Poster 2

The GRID-DP model was proposed to combine data from DInSAR processing with non-radar datasets, which were collected from field study. In this model, first the DInSAR LOS displacement maps generated from unwrapped differential interferograms with a 2-pass DInSAR method was converted into grids of deformation points (DPs) in geodetic coordinate systems. Then, other radar origin datasets of the same area (such as coherence, layover and shadow, slope and aspect, and local incidence angle) and non-radar or external source information (landslide, geology, geomorphology, hydrology etc) required for multi-criteria filtering analysis were similarly extracted in the same size and geographical extent. After merging all extracted grid of points, candidate landslide deformation points (CL-DPs) of high quality radar signals were filtered out by applying GIS-MCFA, which excludes DPs in areas of shadow and overlay, low-coherent, non-detectable and non-landslide deformations, and other possible source of errors from the DInSAR data processing. At the end, the results obtained from GIS-MCFA have been verified by using external datasets collected from fieldworks (existing landslide sites, geological and geomorphologic maps, and rainfall data etc.). 

ABSTRACT

The main goal of this research to detect terrain deformations, movements and subsidence for the determination of hazardous areas to petroleum and gas exploration activities along the Coastal Zones of the Caspian Sea. Differential InSAR was used to determine vulnerable areas to deformations, movements and subsidence along the coastal zones of the Caspian Sea. Multi-temporal 20 ENVISAT ASAR images acquired during 2003-2005 were used for the differential InSAR processing. The following InSAR processing steps were used to detect terrain changes: coregistration, interferogram generation, flat earth removal, noise filter, phase unwrapping, baseline refinement, computation of displacement (line of sight, vertical and horizontal motion).

The contextual information as the historically occurred earthquakes, mud volcanoes and seismic faults were used to validate the achieve results. The hazardous zones in terms of petroleum and gas exploration activities were predicted along the Azerbaijan coastal zones of the Caspian Sea.

 

ACKNOWLEDGMENT

I would like to acknowledge European Space Agency (ESA) for the provision of access to ENVISAT images under the project ID: 18810.

MOTIVATION LETTER

Based on the project proposal to ESA - "InSAR monitoring of oil and gas pipeline corridors in Azerbaijan for detection of terrain movement, deformation and subsidence prone areas and modeling of soil instability (ID: 18810)", Dr. Emil Bayramov was awarded the access to ENVISAT radar images from ESA. According to the commitments to ESA, Dr. Emil Bayramov must present his project results in one of the ESA conferences.

Dr. Emil Bayramov is currently the post-doctoral researcher at Dresden University of Technology in Germany. Dr. Bayramov holds PhD in Natural Sciences from Dresden University of Technology in Germany, MSc in GIS from Lund University in Sweden. Dr. Bayramov was one year fellow at MIT and Harvard University for the study of the geospatial management technologies applied in the urban and regional development. Dr. Bayramov is the Chartered Geographer (GIS) from RGS. Dr. Bayramov holds 15 years of academic and professional experience in GIS and Optical and Radar Remote Sensing.

Revisiting 3D surface displacements in Shanghai, China

with multi-platform SAR images

Shanghai, the largest city in China with a population of more than 24m as of 2013, is located in the Yangtze River Delta in Eastern China. This city has been experiencing ground subsidence since 1920s and the cumulative subsidence was up to 2.63 m in 1965 with an affected area of 400km². A wide range of prevention and control measures has been taken since then by the Shanghai municipal government to mitigate city subsidence, e.g. importing substitute surface water and recharging ground aquifers. Land subsidence was effectively controlled during the period from 1966 to 1995. However, due to the increased water demands, land subsidence at a rate of about 10 mm/year has been observed in recent years. Shanghai Geological Survey Institute reported that the average land subsidence rates were 10.94 mm in 2001 and 10.22 mm in 2002 respectively. This could not only lead to urban flooding and seawater encroachment, but also have significant implications for the safety of urban infrastructure including subway tunnels and gas pipes. It is therefore crucial to monitor ground subsidence in the Shanghai region for predicting potential geological hazards and designing mitigation strategies.

Interferometric Synthetic Aperture Radar (InSAR) has been proven to be a powerful tool for detecting and monitoring city subsidence. Dong et al. applied SBAS InSAR analysis to the ALOS PALSAR datasets from 2007 to 2010, revealing spatial and temporal behavior of land subsidence in a large area of Shanghai(Dong, Samsonov, Yin, Ye, & Cao, 2013). Perissin et al. processed 33 SAR images acquired by the Italian sensor Cosmo-SkyMed with the SARPROZ software. The high density of detected targets make it possible to track the path of newly excavated subway tunnels, allowing us can study the transverse subsidence profile over metro tunnels and predict possible building damage(Perissin, Wang, & Lin, 2012). Chen et al. used Interferometric Point Target Analysis (IPTA) method to extract the mean ground subsidence rates with two different bands of SAR acquisitions, C-band ASAR and X-band CSK SAR. it was found that X-band CSK SAR has a stronger ability to detect non-uniform ground subsidence both in space and time at the urban area (Chen et al., 2013).

Due to the side-looking geometry of SAR systems, conventional InSAR only provides surface displacements in the radar line of sight (LOS). Several researches e.g. (Fialko, Simons, & Agnew, 2001; Gudmundsson et al., 2002; Hu et al., 2014; Wright, Parsons, & Lu, 2004) have demonstrated the possibility to derive 3D displacements from radar observations with various methods: (i) combination of InSAR observations from different geometries, (ii) combination InSAR LOS with SAR azimuth offset measurements; (iii) combination InSAR LOS with Multiple Aperture InSAR (MAI) observations; (iv) integration of InSAR and GPS; and (v) assistance with prior information. It should be noted that, in most cases of city subsidence, the uncertainties in SAR azimuth offsets and MAI observations are greater than the magnitude of horizontal surface displacements. In this study, we attempt to use one ascending COSMO-SkyMED, two descending TerraSAR-X and COSMO-SkyMED SAR images to recover 3D surface displacements in Shanghai.

Each track of SAR images were processed with ultra-short spatial baseline (USB) Temporarily Coherent Points InSAR (USB-TCP-InSAR)(Guoxiang Liu, 2014; Zhang et al., 2012). The TCPs are defined as the points remaining steady for one or several intervals of SAR acquisitions. It can detect coherent points from a few images and acquire reliable solution without phase unwrapping(Lei Zhang, 2011). The successful application on extracting deformation on Los Angeles Basin with ERS-1/2 images indicates that TCPInSAR is a promising alternative for efficiently mapping ground deformation even from a relatively smaller set of interferograms(Zhang et al., 2012). As the topographic effect is not sensitive in the interferograms with very short baseline, an external digital elevation model (DEM) is no longer needed in interferometric processing, and the parameters needed to be estimated were simplified at the same time. The high-rise buildings in Shanghai and the absence of DEM would test the feasibility of USB-TCPInSAR approach.

The results showed that USB-TCPInSAR can detect deformation effectively using only one set of SAR images(less than 15 images). A big funnel-shaped subsidence holes with a speed of -20 mm/year around Hongkou stadium was found for all data sets, which is consistent to previous work. It was found that for the same satellite data obtained in ascending/descending orbits, the difference between them reached to a maximum of 8mm/year. While for the different satellite data obtained in same descending orbits, the result showed great consistency, which proved that the horizontal motion cannot be ignored even on the urban area with very slow horizontal motion. The derived 3D vertical displacement is quite different to the three individual LOS direction. Comparison between leveling points and USB-TCPInSAR results were perform, showing that the three-dimensional vertical deformation was closest to the leveling results with a standard deviation of 0.57 mm/year. This research indicates that with the increasingly high accuracy demand, the small horizontal motion on urban area must be taken into account to estimate the accurate vertical deformation.  

 

References:

 

Chen, J., Wu, J., Zhang, L., Zou, J., Liu, G., Zhang, R., & Yu, B. (2013). Deformation Trend Extraction Based on Multi-Temporal InSAR in Shanghai. Remote Sensing, 5(4), 1774-1786. doi: 10.3390/rs5041774

Dong, S., Samsonov, S., Yin, H., Ye, S., & Cao, Y. (2013). Time-series analysis of subsidence associated with rapid urbanization in Shanghai, China measured with SBAS InSAR method. Environmental Earth Sciences. doi: 10.1007/s12665-013-2990-y

Fialko, Y., Simons, M., & Agnew, D. (2001). The complete (3-D) surface displacement field in the epicentral area of the 1999 M(w)7.1 Hector Mine earthquake, California, from space geodetic observations. Geophysical Research Letters, 28(16), 3063-3066. doi: Doi 10.1029/2001gl013174

Gudmundsson, S., Gudmundsson, M. T., Bjornsson, H., Sigmundsson, F., Rott, H., & Carstensen, J. M. (2002). Three-dimensional glacier surface motion maps at the Gjalp eruption site, Iceland, inferred from combining InSAR and other ice-displacement data. Annals Of Glaciology, Vol 34, 2002, 34, 315-322. doi: Doi 10.3189/172756402781817833

Guoxiang Liu, H. J., Yunju Nie, Tao Li, Rui Zhang, Bing Yu, Zhilin Li. (2014). Detecting Subsidence in Coastal Areas by Ultrashort-Baseline TCPInSAR on the Time Series of High-Resolution TerraSAR-X Images. IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 52(4).

Hu, J., Li, Z. W., Ding, X. L., Zhu, J. J., Zhang, L., & Sun, Q. (2014). Resolving three-dimensional surface displacements from InSAR measurements: A review. Earth-Science Reviews, 133, 1-17. doi: 10.1016/j.earscirev.2014.02.005

Lei Zhang, X. D., Zhong Lu. (2011). Modeling PSInSAR Time Series without phase unwrapping. IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 49(1).

Perissin, D., Wang, Z., & Lin, H. (2012). Shanghai subway tunnels and highways monitoring through Cosmo-SkyMed Persistent Scatterers. ISPRS Journal of Photogrammetry and Remote Sensing, 73, 58-67. doi: 10.1016/j.isprsjprs.2012.07.002

Wright, T. J., Parsons, B. E., & Lu, Z. (2004). Toward mapping surface deformation in three dimensions using InSAR. Geophysical Research Letters, 31(1). doi: Artn L01607 Doi 10.1029/2003gl018827

Zhang, L., Lu, Z., Ding, X., Jung, H.-s., Feng, G., & Lee, C.-W. (2012). Mapping ground surface deformation using temporarily coherent point SAR interferometry: Application to Los Angeles Basin. Remote Sensing of Environment, 117, 429-439. doi: 10.1016/j.rse.2011.10.020

 

An intensive oil production is known to be able to result in increased pressure on the subsurface which, in turn, leads to the rise of seismic activity (manifested in the form of earthquakes, for instance) and vertical strains, and vertical movements of the ground. These deformations may cause various accidents that can harm the environment and violate the regime of oil-production enterprises. Thus, monitoring of the earth surface deformation processes within regions with intense oil production becomes today quite important.

Such monitoring is generally carried out by traditional geodetic techniques, such as high-precision leveling or GPS positioning. However, these methods are very expensive and time-consuming, while monitoring using certain radar data gives an advantage over the classical methods in much more large coverage of the area and in speed of information processing.

 Despite the rapid development of the method, in Russia such studies are still extremely rare and, in most cases, are experimental. Therefore, the main objective of this work was to estimate the informativity of the remote sensing for monitoring and evaluation of modern crustal movements within the oil-producing regions of the Republic of Tatarstan that includes one of the biggest Russian oilfields (reserves are estimated at 5 billion tones).

 In this research Differential Interferometry technique was applied to data obtained from two satellites: ERS-2 (for reconnaissance) and ALOS PALSAR (for more accurate observations).

 Each pixel of radar image represents the coherent sum of the signal reflected from the point targets located on the surface within the resolution cell. Comparison of coherence maps with geographical features of the area shows that the best targets for monitoring using radar sensing are urbanized areas. In such areas one can see the highest density of pixels with high coherence. Therefore, for these areas one can build the most accurate and reliable displacement maps.

 From this point of view, the most interesting object is the territory of Almetyevsk city. Citizens have repeatedly experienced the effects of earthquakes, and there are hundreds of wells located around the city. It would seem that the city territory should undergo a systematic subsidence, but somewhat different results were obtained.

 A series of displacement maps obtained for this study shows that the territory of Almetyevsk experienced predominantly upward vertical movements, with the most intense uplift along a linear zone approximately 1.5 km wide. It is also interesting to note that the displacements observed are not of linear character. Perhaps these movements have a periodic nature (the average speed of vertical movements is about 5 mm per year). Unfortunately lack of data does not allow evaluating the period and giving a reasonable interpretation of this phenomenon.

 Nevertheless, the current study results show prospectivity of such monitoring, and ERS data can be used for reconnaissance, while for more accurate observations higher resolution systems with bigger wavelength should be used (ALOS PALSAR, for instance).

 Reliability of the data obtained by radar interferometry, as a rule, verifies by high-precision geodetic leveling. To compare displacement magnitudes obtained by interferometry and leveling graphs of vertical displacements were also plotted. The results show that accuracy of the former is equal to accuracy of the latter while the processing speed is much more high. It is also clear enough that resolution of the interferometry technique is much greater than resolution of the traditional method. Little differences in absolute values can be explained by inaccuracies of both methods and inaccuracies occurred during the process of positioning leveling chains relative to radar images.

In conclusion, considering the fact that availability of remote sensing data increases rapidly nowadays one can believe that monitoring of earth surface displacements from space in the near future will become one of the key technologies of prevention and prediction of man-made and natural disasters.

Tags: remote sensing, DInSAR, monitoring of land subsidence, oil-producing regions

From direct consumption and food preservation to its impact on hydrocarbon accumulation, rock salt is an invaluable resource because of its characteristically low geothermal gradient, density, and viscosity. But as globally inimitable and functional as salt is, our primary form of observing salt is via subsurface imaging techniques because it is so rarely preserved on Earth’s surface. Only few examples of surface salt bodies exist today despite countless interpretations of subsurface salt bodies throughout the world that indicate subaerial exposure during its structural evolution. Subaerial salt exposures tend to be transient periods during the evolution of a salt body but its impact on the salts’ subsequent structural development can be profound. The study of surface salt kinematics is critical because the shape a salt body develops as it flows on the surface prior to sedimentary burial strongly impacts the kinematic evolution of its overburden and controls the subsequent salt distribution. Much of current knowledge of surface salt kinematics stems from surface salt glacier, or namakier (following Baikpour & Talbot, 2010), studies predominantly in the Alborz and Zagros Mountains in north and south Iran, respectively. However, recently recognized namakiers in western Kuqa fold thrust belt, northwestern China provide a fresh, natural laboratory to observe and constrain subaerial salt kinematics. A recent field study by Li et al., [2014] in western Kuqa fold thrust belt indicate active deformation of namakiers based from stratigraphic and morphological relationships. Satellite geodetic measurements confirm that namakiers in Kuqa fold thrust belt are active.

The Kuqa foreland fold thrust belt actively accommodates north-south shortening between the Tian Shan to the north and the Tarim basin to the south across an approximately 400 km long, and 20 - 65 km wide region. The Paleocene-Eocene Kumugeliemu Group in western Kuqa fold thrust belt outcrops at several locations that include line sourced (Tuzimazha salt wall, Quele salt thrust and its associated salt namakier) and point sourced (Awate namakier, Bozidun namakier and an unspecified structure referred to hereafter as the Western namakier). We focus on the most active line sourced and point sourced namakiers in Kuqa foreland fold thrust belt, Quele and Western namakier. The Quele salt thrust propagates southward, just south of the Baicheng synclinal basin, and the salt décollement is exhumed where the thrust fault breeches the surface, forming the Quele namakier. The Quele namakier extends ~35 km E-W with a lateral thickness ranging from ~50 m to ~200 m. The Western namakier is about 4.5 km long and 3.4 km wide and is interpreted to be a point sourced structure based on its surface geometry. Using active surface displacement measurements from interferometry of synthetic aperture radar (InSAR) we use numerical models to understand the mechanical properties of namakiers in western Kuqa fold thrust belt of Xinjiang Province, China. The InSAR velocity estimates were calculated from 40 Envisat ASAR C band images between June 2003 and October 2010 from descending track 291. InSAR processing was done with the New Small Baseline Algorithm Subset (NSBAS) chain as described by Doin et al., 2011. 

The Black Sea coast of the Great Caucasus has always been a region of high landslide risk due to widely spread clays and marls seasonally saturated by abundant rainfalls. In recent years landslide risk assessment has become vital because of strongly increased human-induced impact dealt with construction of the Sochi-2014 Olympic Games facilities. We studied the landslide activity in the area of the Bolshoy Sochi (Big Sochi) using the StaMPS PS-InSAR and SBAS methods. We incorporated all available radar images from the satellites with different wavelengths: 17 ALOS images (track 588A, ascending orbit, 22.01.2007-17.09.2010), 13 Envisat images (track 35D, descending orbit, 29.11.2010-23.03.2012), 17 TerraSAR-X images (track 54A, ascending orbit, 24.12.2011-13.09.2012). For all of these three sets of images we identified persistent scatterers (PS) with a density of more than 350 PS/km even for rural areas. For these PS’s we calculated surface displacements in the line of site (LOS) direction. Areas of relatively high surface displacement rates located using the SAR data appeared to match in many instances zones of high landslide activity identified by geological field observations. We present time series for the most active landslide areas of the Big Sochi. The calculated average surface LOS displacement rates vary from 15-20 mm/year up to 60 mm/year. For some places with extreme values of LOS displacement rates the values of displacement rates in the down - slope direction were also estimated. 

Analysis of the time series made it possible to determine periods of activity and relative stability of the landslides and compare them with ground observations. For example, high landslide activity in spring of 2012 initiated by heavy rainfalls can be clearly seen in the time series. Activity of the landslides in the beginning of 2012 is most pronounced in the results of PS-InSAR processing of the TerraSAR-X data. Detailed analysis of the time series is performed for the Baranovka landslide where very strong movements during 23-24 January 2012 damaged 35 houses. Large displacements resulted in temporal decorrelation between the SAR image from 24.12.2011 and the later images. Reliable estimation of displacement rate appeared to be possible only since 17.02.2012. There have been revealed the period of relatively high landslide activity (17.02.2012 - 10.03.2012) when the LOS displacement rates reached 30 mm/month (about 120 mm/month down the slope). Then displacement rates decreased gradually and since June 6, 2012 have not exceed 2-3 mm/month (10 мм/month down the slope), indicating that the landslide has become more stable. Displacement fields for the Baranovka landslide for the period of large displacements were also estimated using pairs of interferograms. These estimates are in good agreement with time series obtained for the corresponding PS’s. Results obtained with the StaMPS software are compared to some of those using the DePSI and GiANT (SBAS, NSBAS) methods. DePSI produces nearly the same values of displacement rates as StaMPS for the Mamaika landslide. SBAS makes it possible to locate the area of relatively strong movements on the Mamaika landslide but calculated values of displacement rates are lower than using StaMPS. All this illustrates high efficiency of  SAR Interferometry in monitoring of landslide activity in the area of the Black Sea coast of the Caucasus where landslides are numerous.

Abstract

There are more than 250 salt domes in south of Iran and Persian Gulf which belong to a unit sedimentary basin Hormuz salt and separated by further tectonic factors. Salt dome outcrop in Zagros orogenic Simply Folded Belt (SFB) is a unique composition of exotic rocks which reached the earth surface from depth of 3-5 km by Hormuz evaporates in Late Precambrian. Estimates of the total stratigraphic thickness vary from 10 to 15 km and detached from the 1–2 km thick infra-Cambrian Hormuz. The Hormuz series penetrates the overburden as diapirs that feed spectacular salt glaciers moving by plastic flow along the ground surface. Most of salt diapers of Zagros have close relationship with the regional geodynamic history as well as the hot arid of Persian Gulf can control the soluble salt escape rapid erosion.

We investigate the surface deformation of Anguru (Gasho) salt diapir at the west of Bandar-Abbas (27^o17' N, 55^o51' E) by advanced Interferometric Synthetic Aperture Radar (InSAR). The rates of flow of salt within diapirs have traditionally been measured by standard survey techniques in the previous works while InSAR can provide a quick review of the Quaternary motions of most observed surface salt flows in some elevated diapirs which are impossible in-situ measurements. Barnhart and Lohman (2010)identified Anguru as an active salt that showing good coherence in individual ALOS and Envisat interferograms. Due to high surface velocities or steep local topography performing classical InSAR time-series analysis is challenging at Anguru. We performed advanced analysis and modified several aspects within the processing chain using Small BAseline Subset (MSBAS) method to create time-dependent deformation signal using ALOS data.The modification includes filtering prior to phase unwrapping, topographic correction within 3-dimentioanl phase unwrapping and orbital correction to estimate deformation. By this modification, we were able to obtain 1421 coherent pixels over Anguru and subsidence deformation continuously observes in time series. The signature of different selective points within deforming region show approximately same behavior and suggest that major part of salt diapir broadly sustains subsidence, although the local uplift is also recorded.

 

Key words: Hormuz salt, Zagros, Anguru, MSBAS algorithm

In this study we compare TerraSAR-X (TSX) Staring SpotLight (ST) Mode data with Sentinel-1A (S1) Interferometric Wide (IW) swath mode data for landslide movement monitoring. Differences in ground resolution, revisit time, number of achievable measurement pixels, coherence behavior, maximum detectable movement and applicable interferometric approaches shall be discussed for these two kinds of synthetic aperture radar (SAR) data.

 

TSX operates in the X-band with a wavelength of 3.1 cm and a revisit cycle of 11 days. Its ST mode allows for a spatial resolution of 0.24 m x 0.6 m and scene size of around 4km (width) x 3.7km (length). S1 has a 12-day repeat cycle and carries a single C-band synthetic aperture radar instrument with a wavelength of about 5.5 cm. Its pre-defined mode over land is the IW mode with a 250 km swath width and moderate 5 m x 20 m spatial resolution. The interferometric phase as a result of radar phase signal processing allows the derivation of surface movement values, beside other estimation. Exploiting a stack of radar data the spatio-temporal behavior of surface movements can be analyzed in the meaning of time series analysis. 

 

The main objective is to measure the movement of a small scale landslide close to urban infrastructure in Colombia. The area of interest covers a city adjacent to the landslide that is approximately ~1 km x ~1 km in size. Urban infrastructure characterizes the area of interest and light vegetation covers predominantly the land slide. It has been recently tried to develop this area of landslide for urban use (prospering areas), mainly to build houses for inhabitants. Despite the installation of stabilization constructions and due to the strong movement of the landslide several new built houses have been destroyed. The remains and small scale man made features, like concrete structures to stabilize the land slide itself, are the available radar scatterers within the area.

 

Theoretically, it can be expected that the very high TSX ST spatial resolution will provide more measurement pixels than the coarser S1 resolution over this spatially limited area of interest. Consequently, TSX may show more detail than S1 for the given landslide. To investigate or monitor the movement of the land slide the number of achievable measurement pixels to be analyzed is an important number. Since the concept of interferometric remote sensing makes only sense if sufficient measurement pixels are available.

 

Both satellites have a comparable revisit time interval of 11 and 12 days, respectively. C-band satellite data, as S1 does acquire, are known for lower temporal decorrelation behavior than X-band data. A modern interferometric approach, capable to exploit point scaterers and distributed scatterers, may benefit from this coherence behavior. Beside the number of achievable measurement pixels the temporal decorrelation is essential parameter for meaningful interferometric remote sensing of surface movements.

 

Regarding the maximum movement between two neighboring measurement pixels within one revisit, TSX allows for a maximum of 0.7 cm unambiguous movement in line-of-sight. The equivalent value for S1 is 1.3 cm. That is, S1 could be better suited to map faster moving landslides than TSX.    

 

Investigating theoretical and practical differences between the selected TSX ST and S1 IW data regarding surface movement monitoring of small scale landslide adjacent to urban infrastructure will lead to a recommendation of more suited data type.

The poster presents first results of the project “The integrated system of surface deformation monitoring based on Persistent Scatterer Interferometry, measurements from permanent GNSS stations and precise leveling” (Cat-1 ESA project ID 19409).

The main aim of the project is to create integrated system for monitoring ground deformations caused by man-made factors, based on radar satellite interferometry (PSI), permanent GNSS stations and precise levelling.  Within this topic the main methodical objectives aim at preparation of new algorithms, which will integrate results of interferometric observations applying various angles and polarizations with the measurements obtained from two remaining methods, as well as cognitive goals concerning relations between geological base and construction works will be achieved.

Warsawagglomeration has been selected as a study area; it is characterized by different ground deformations (geological-based large-area deformations and local-based man-made deformations) and by complex geological structures. Recent local ground deformations inWarsawcaused by construction works (e.g. construction of the second metro line, large buildings and skyscrapers) are analyzed using high-resolution CosmoSkymed (CSM) radar images collected between 2011 – 2016. History of vertical movements of earth surface in Warsaw agglomeration in the period 1992 – 2010 are re-created on the basis of analysis of archival radar images collected by ERS-1/2 and ENVISAT satellites (various polarizations and viewing angles) It will enable to understand better and interpret local deformations observed on high-resolution data in spatial and temporal context. The prepared methods of joint analysis of ENVISAT images collected with various angles and polarizations will be used for analysis of Sentinel-1 images, which are planned to be acquired starting from 2014.

Results of interferometric analyses are validated with the use of large archival database of levelling measurements and present measurements (including long leveling traverses between permanent GNSS stations). Results gathered from three techniques are analyzed jointly, referring to complex geological structure and man-made movements; radar interferometry integrated with the remaining techniques will be used for preparation of hazard maps. It will help to foresee and evaluate effects of conducting construction works in particular geological structures.

Temporal monitoring of terrain movements and satellite observations are used here to monitor geophysical Natural Hazards in northern Greece, in an attempt to address their causes: anthropogenic or natural due to geological and geomorphological evolution. Site-specific preliminary results are here presented for the broader urban area of Thessaloniki. It is the second most populated city in Greece with many important industrial constructions as well as life-lines situated in its proximity, making the hazard evaluation imperative. From a tectonic point of view the city is surrounded by normal and strike-slip faulting, striking roughly E-W. One of these normal faults moved during the June 1978 Mw6.5 earthquake, which caused significant damage to the city and loss of life.

The well-established Persistent Scatterer Interferometry (PSI) and the Small Baseline Subset (SBAS) of Interferometric Synthetic Aperture Radar (InSAR) techniques are applied to process 46 ERS 1/2 satellite and 35 ENVISAT images, that cover the region of study, from 1992 to 2010.  For the analysis, more than 250 interferograms were formed, using raw data provided from the European Space Agency (ESA). The process aims to detect small-scale movements, tectonic movements and movements triggered by the overexploitation of the aquifers. The broader area of interest consists primarily of agricultural lands, leading to limited interferometric coherence, and therefore careful interferogram filtering and phase unwrapping of the input stack was crucial for reliable velocity generation. Final velocities are integrated with geodetic GPS measurements, and for the final interpretation geological, hydrogeological and geotechnical data are also considered.

The satellite imagery revealed two significantly deforming areas: Kalochori, in the west and Anthemountas Basin, in the east. Both of these areas are related with human activities and residential constructions. The wider Kalochori region, related with the industrial activity of the broader area of Thessaloniki, is located below sea level suffering from land subsidence for more than 50 years, with several marine invasions reported in the past. Moreover, the impact of the economic crisis on the industry sector led to a different water consumption.   

The ground deformation of Anthemountas Basin is controlled by the potentially active and active surrounding faults, and different deformation characteristics are detected along these faults systems. The ground deformation of the basin, which also hosts the “Macedonia Airport” an International airport used by around 4 million travelers every year, is crucial for the economic and human activities in northern Greece.

Finally, the products are contributing to the seamless monitoring of the region, which acts as a junction of the Balkans, especially important in the new Sentinel era where shorter revisit times will potentially provide systematic geodetic measurements over the area.  

This work was partially funded by a) the European Union 7^th Framework Programme (FP7-REGPOT-2012-2013-1), in the framework of the project BEYOND “Building a Centre of Excellence for EO-based monitoring of Natural Disasters”, G. A. 316210 and b) co-financed by the European Union (European Social Fund—ESF) and Greek national funds through the Operational Program “Education and Lifelong Learning” of the National Strategic Reference Framework (NSRF)—Research Funding Program: Thales - Investing in knowledge society through the European Social Fund (MIS377335). 

The excessive groundwater extraction and the consequent overexploitation of the aquifers to satisfy the increasing water demand in many Mexican cities have led to ground subsidence and associated fracturing.  The metropolitan area of the Valley of Mexico (ZMVM), for example, has been severely affected by these phenomena. Several multidisciplinary studies have mainly been concentrated on the surveying, analysis and understanding of the subsidence in the southern ZMVM (e.g., Osmanoglu et al. 2010; Lopez-Quiroz et al. 2009; Mendez et al. 2008; Arango-Galván et al. 2011; Cerca et al. 2012). However, a few related studies have been performed in the northern ZMVM (e.g., Huizar et al. 2003, Siles et al., submitted paper), where important hydraulic projects are found or under development (e.g., Gran Canal del Desagüe, Tunel Emisor Oriente) and others are planned (e.g., artificial water recharge; GDF et al. 2012). This area, like the southern ZMVM, has been experiencing an accelerated demographic growth and an uncontrolled urban sprawl over potential natural recharge zones for the local aquifer and areas prone to geological hazards (either natural- or human-induced). Thus, precise subsidence mapping, definition of fracturing risk zones and their continuous surveillance over this area is essential to mitigate their effects. Radar remote sensing techniques, such as multi-pass InSAR, can enable temporal and spatial coverage of ground deformation, providing valuable information to accomplish these tasks. In this work, a detailed ERS and ENVISAT time series deformation analysis on northern ZMVM is presented, covering a period of approximately 11 years (1999-2010). The small baseline (SBAS) method from the Stanford method for Persistent Scatterers (Stamps) was used for that purpose, being adapted for selection of an appropriate reference area based on the surface geology data. Population growth and water extraction rates were also considered for the analysis of the ground sinking development in this area. Our results indicate increasing subsidence rates for the period of study reaching up to 285 mm/yr. Land settlement in the area was observed to decrease in south-north direction and towards the highlands that surround the lacustrine plain of primarily compressible deposits; this distribution approximately coinciding with that of the clay-deposits thickness. Both urban and agriculture areas have been identified to be subsiding, being more notable in the former case. Although subsidence essentially exhibits a linear behaviour, a non-linear component was also identified, evaluated and modelled. This component, presumably suggesting local deformation anomalies, could be related to interaction of complex earth processes occurring in the area such as fracturing, tectonic activity and different soil consolidation stages. Moreover, impact of subsidence and associated events on important local infrastructure is described, and velocity gradients are used to locate probable fractures in the area. Triggering fracturing mechanisms are addressed by taking into account the available geological and groundwater pumping information as well as our InSAR outcomes. Most cracking activity in the area may be caused by differential subsidence; nevertheless, other factors such as buried geological structures, water infiltration into existing discontinuities and seismic activity could also play an important role in the generation and propagation of fractures. The present work intends to provide useful and valuable information for improvement and assessment of vulnerability maps in order to support decision-making processes in the area of study. Additionally, the detection of fracturing zones by the aid of satellite data (enhanced with in-situ measurements) can also contribute to both waste and water management by proving critical spatial information for proper location of deposal sites and water treatment plants.

Engineered earthen covers are used at land disposal sites to prevent or minimize releases of a variety of hazardous substances to the environment, such as radionuclides and hazardous organic and inorganic chemicals. The performance of the covers can be degraded by generalized erosion and gully formation.  Standard methods of monitoring consist of visual inspection and may include elevation surveys of a few locations where tailings are settling, but site‑wide elevation surveys to detect areas of ponding or incipient gully formation are typically not performed.

This research investigated the feasibility of using complementary satellite‑based techniques, such as coherence analysis, differential radar interferometry (DInSAR) and multitemporal interferometry (in particular the small baseline subset (SBAS) technique), to monitor the performance and stability of soil covers at uranium mill tailings sites and other land disposal sites with earthen covers. The research also advanced the understanding of the effect of rip rap on the synthetic aperture radar (SAR) signal response.

Archived ERS-1/-2 SAR data were acquired to include two uranium mill tailings sites in northwestern New Mexico: the Bluewater site and the Grants Homestake site. The archived data covered the period 1995 to 2002. These SAR data sets were screened based on meteorological conditions at the time of acquisition, to develop a subset of SAR data that were most likely to yield good results. Initially, data were used to investigate the strength and coherence of the radar signal from actual mill tailings covers.  Mill tailings covers may consist of vegetated soil, rock mulch (a mixture of rock and soil), or rip rap (coarse rock).  These various cover conditions affect the strength of the SAR signal and may affect its coherence, which in turn affects the ability to use InSAR techniques in estimating surface displacement measurements. 

DInSAR requires that SAR data preserve coherence, which in case of Bluewater and Grants sites should be acquired no more than few months apart. For the Bluewater site, the DInSAR and SBAS analyses provided consistent centimeter-scale changes in ground surface displacements, mapped over the entire tailings impoundment. The displacements correlate with variations in the slope of the tailings impoundment cover and nature of the underlying tailings. At the Grants site, only the rock‑covered side slopes maintained sufficient coherence to perform quantitative SBAS InSAR analysis, because reclamation activities disturbed the surface soil of the impoundment cap during the time period covered by the SAR data.  The InSAR analysis indicated the south side slope showed downward movement with respect to the other side slope.

 

Overall, this research provided a proof-of-concept that satellite based techniques techniques are useful for monitoring the performance and stability of soil covers at uranium mill tailings sites and other land disposal sites with earthen covers.  Coherence analysis provides a fast qualitative way to detect stable areas that remain coherent over a long period of time. InSAR analyses provided a way to quantitatively assess rates of change for unstable areas. The extent of the analysis that could be performed for the two locations was limited by the availability of archived SAR data, but the results demonstrate that future monitoring using on-demand SAR data acquisitions can yield reliable measurements of ground surface displacements to guide maintenance of the cover. The work also is relevant to land disposal of other types of non-high-level nuclear waste, uranium waste rock disposal, and mine reclamation. 

Nyiragongo and Nyamulagira are two of the most active African volcanoes. They are located in the Virunga Volcanic Province, in Democratic Republic of Congo, in the Western branch of the East African Rift.

 Nyiragongo hosts a ~200 m wide semi-permanent lava lake in its 1.2 km wide crater. The lava lake level fluctuates significantly in its pit. It can drop down of few tens of meters below the bottom of the crater. Sometimes, lave rises enough to overflow the surrounding platform.  After solidifying, these successive overflows make the bottom of the crater to rise up to few tens of meters per year. Except for a temporary prototype of Stereographic Time Lapse camera (Smets et al. in prep), no monitoring system exists to measure the changes occurring in the crater: lava lake level fluctuation, filling of the crater by successive lava overflows, changes in the dimensions of the lava lake and deformations of the bottom platform.

Nyamulagira also used to host a lava lake in a pit located in the northeastern sector of its 2,2 km wide caldera. It was however drained during a voluminous eruption accompanied by the partial collapse of its crater in 1938. Since then, activity at Nyamulagira is characterized by frequent flank eruptions sometimes accompanied by fumarolic emissions and small lava flows within the caldera. However, since the last voluminous eruption that took place 12km northeast of the crater in 2011-2012, intense permanent degassing and long period seismicity suggested that lava must be emplaced close to the surface somewhere in the caldera.  After two years, lava fountains were finally observed in the pit crater on June 2014. Field observations allowed estimating that the dimensions of the pit crater had dramatically evolved since the 2011-2012 eruption, changing from a ~50 m deep and 400 m wide pit into a 500 m deep and 400 x 600m wide crater (Smets et al., 2014). That recent activity suggests that a new lava lake may develop again in the Nyamulagira crater. However, because of the insecurity affecting that region, as for Nyiragongo, no ground-based monitoring system allows measuring the activity and the changes occurring in the crater.

 We will show how the use of high-resolution Strip Map TerraSAR-X and TanDEM-X images (<3 m ground resolution) and Ultra Fine Radarsat-2 C-band images (1x2m resolution) allows studying the current activity in these two active craters. We make use of classical DinSAR, InSAR SBAS time series methods as well as advanced Split Band Interferometry (Derauw et al., this meeting) for monitoring the changes and computing high resolution Digital Elevation Models.

 Advanced MSBAS time series method (Samsonov and d’Oreye, 2012) is also applied to the large SAR database (ERS, ENVISAT, ALOS) acquired in the frame of successive projects (Kervyn et al., this meeting), allowing comparing the present activity with the one that occurred over the last decade.

 Whenever possible these data are compared with ground-based 3D models of Nyiragongo crater computed using the Stereographic Time Lapse camera system that operated few months from September 2011.

 Very high-resolution SAR images and advanced InSAR methods offer unique opportunities to simultaneously study these two neighboring active craters separated by only 13 km.

 References:

 Smets B., d’Oreye N. ,and Kervyn F., Toward Another Lava Lake in the Virunga Volcanic Field?. Eos Trans. AGU, Vol 95, No 42, pp377-378, 21 October 2014

10.1002/2014EO420001

 Samsonov S. and N. d’Oreye, Multidimensional time series analysis of ground deformation from multiple InSAR data sets applied to Virunga Volcanic Province. Geophysical Journal International, vol. 191, Iss. 3, 1095-1108, 2012

The mining exploitation of hard coal in Upper Silesia Coal Basin (USCB) - Southern Poland has been conducted since XIX century and continues to the present days. Nowadays USCB is one of the largest hard coal mining areas in Europe. It is located in Southern Poland and covers an area of over 6000 km². During the period 1945-1979, on average, 200 million tonnes of hard coal were exploited yearly. From the late 80s, a steady decrease reached 79 million tonnes in 2012. Urban growth was parallel to mining industry development, reaching 37 towns with almost 3 million inhabitants nowadays. The most common operating system in USCB is longwall coal mining. Usually, the excavated coal layer is 2,5 m thick, 250 to 400 m long, and about 680 m deep. According to  subsidence can reach up to 70% of the excavated coal layer, which represents a 0,75-2,0 m displacement for every layer. Different methods of monitoring subsidence of the earth surface were applied in this area, starting from analysis of topographic maps compiled in different period of time, precise leveling, laser scanning and GPS measurements. Nowadays satellite interferometry seems to be useful and effective tool for these purposes. DInSAR technology was used to identification and monitoring subsidence basins caused by underground coal mining activities. Two sets of differential interferograms were used for comparison and analysis: ALOS – PALSAR, period 22/02/2007 – 27/05/2008 and TerraSAR‑X, period 05/07/2011- 21/06/2012. The analysis has covered the whole territory of the Polish Upper Silesia and small part of the Czech coal basin in the Ostrava – Karvina region. The analysis of the ALOS – PALSAR data enabled identification of 112 subsidence basins. Most of the subsidence basins were located within mining area, however in some cases they exceeded boundaries of mining areas. In several cases it was also possible to determine direction and development of subsidence movement which  correlated most probably with movement of underground mining fronts. Identified subsidence basins are arranged in about longitudinal zone in central part of the Upper Silesian Basin almost continuously from Mysłowice  to Rybnik. In addition, there are two other zones: in the north, latitudinal between Piekary Śląskie and Bytom and in the south WNW – ESE oriented between Jastrzębie Zdrój and Wodzisław. The comparison of subsidence basin was performed between years 2008 and 2012. It was made in a northern and central part of USCB between Piekary Śląskie (N), Zabrze (W), Tychy (S) and Mysłowice (E). It is clearly visible that the number and size of the subsidence basins is smaller in 2012 – 31 in comparison to 51 in 2008. Data obtained from Register of Mineral Resources and Mining Areas Information led by Polish Geological Institute show that in both periods of time hard coal extraction was carried out with similar amount of deposits 28 in 2008 and 27 in 2012. Although total extraction at the area which was subject of comparison was 39 million tonnes in 2008 and 31 million tonnes in 2012, so lower after four years by 8 million tones. These data can indicate that decreasing of the subsidence basins number is related to the reduction of the hard coal exploitation. On the other hand interferometric data which were compared, are characterized by different  wavelength, ALOS – PALSAR 25 cm (lower resolution) and TerraSAR – X 3.1 cm. It could have some impact on the interpretation and identification of the “fringes” indicating the subsidence basins. The obtained results emphasize the importance of satellite interferometry method for monitoring land surface at the mining areas and should be used for verification of the local territorial plans and should be taken under consideration during risks assessment related to mining activity.

Sinkholes are circular depressions, ranging in width and depth from less than a meter to hundreds of meters, that form by a sudden collapse of the surface. Sinkholes occur in a diversity of geological environments around the globe. Their most common formation mechanism is dissolution of soluble rocks and creation of subsurface cavities that collapse after the overlying layers become mechanically unsupported. Sinkholes along the Dead Sea shorelines (Israel and Jordan) are attributed to cavities that formed by dissolution of a subsurface salt layer, up to 20 m thick and 50 m deep. This dissolution occurs due to the replacement of hypersaline groundwater by fresh groundwater as the Dead Sea level and the fresh-saline water interface drops and recedes. The first identified sinkholes along the Dead Sea coast appeared in the 1980’s. Their occurrence rate has accelerated significantly from less than 50 new sinkholes a year before 1999 to more than 400 sinkholes a year since 2003, reaching the current rate of about 450 new sinkholes a year. This increasing growth rate is correlated with the increasing rate of the Dead Sea level drop, which is currently more than 1 m per year.

Sinkholes along the Dead Sea shorelines are predated by, associated with, and followed by gradual land subsidence, which continues for months to years after the sinkholes collapse. While the surface processes around and within the sinkholes are relatively well resolved, the underlying, subsurface formation mechanisms are less understood and poorly resolved. Of particular interest are basic topics such as the volumes involved in the dissolution and collapse processes, the cavities growth patterns, and the specific spatial and temporal relations between gradual subsidence and sinkholes. To resolve these issues we measure the ground subsidence around sinkholes from its onset to the collapse of the sinkhole/s, at three test areas in Israel using Interferometric Synthetic Aperture Radar (InSAR). The measurements are carried out with COSMO-SkyMed images processed at full resolution of about 3 m, using a high resolution (0.5 m) digital elevation model derived from airborne LiDAR measurements. We resolve the shapes and dimensions of the layers that bend into the cavities by elastic inverse modeling of the observed ground subsidence. The variable bending into the cavity roof is simulated as a planar array of contracting cracks ("anti-sills"). The bending volume inferred from modeling, added to the collapsed sinkhole volume measured in-situ, provide an estimate of the total volume of salt layer dissolution. A Coulomb Failure Function (CFF) analysis is then applied to calculate the stress field induced by the bending layers, at 1 month intervals starting from the onset of subsidence until the occurrence of the first collapse. We find that the induced stress field favors the generation of sinkholes and fractures at the perimeters of the subsiding areas, in agreement with field observations, rather than at their centers (as intuitively expected). Finally, our analysis enables to trace lateral migration of cavities along with the salt dissolution front, implying that previously suggested models of vertical cavity migration are less likely.

To overcome the shortage of space, the Hong Kong government has been reclaiming lands along the seafront, which will introduce continuous ground settlement since the beginning of reclamation. The Hong Kong International Airport, 75% of which was reclaimed from the sea, is one of the largest artificial islands in the world (Jiang & Lin, 2010). Figure 1 presents the fill allocation in the airport platform. Chek Lap Kok and Lam Chau (the white color regions in Figure 1) are two original islands located in the sea. The surrounding areas are reclaimed using different fill types.

Since the beginning of reclamation, the ground and facilities may be subject to various movements caused by unconsolidated subsurface soil layers, penetration of seawater, tidal load, the impacting force of the landing airplanes, and etc. The revelation of its dynamic process and driving forces requires multidisciplinary knowledge of geology, hydrology, meteorology, structural engineering, and etc. To this end, this paper jointly exploits L-band ALOS/PalSAR, C-band ENVISAT/ASAR, and X-band TerraSAR-X data to monitor the deformation of the airport from 2007 to 2011. The multi-dimensional SAR tomography technique (Fornaro, Reale, & Serafino, 2009; Lombardini, 2005; Xiao Xiang & Bamler, 2010) is employed to extract linear deformation velocity of single persistent scatterers (PSs). The results are shown in Figure 2. The deformation results are evaluated using leveling survey data of benchmarks obtained from the Hong Kong Airport Authority and height results are validated using LiDAR data obtained from the Civil Engineering and Development Department. We can observe that the number of detected PSs using high resolution TerraSAR-X data increases dramatically compared with that using medium resolution ALOS/PalSAR and ENVISAT/ASAR data. The derived linear deformation velocity maps using ALOS/PalSAR, ENVISAT/ASAR, and TerraSAR-X data show a good agreement in spatial dimension notwithstanding different looking angles. The result using TerraSAR-X data confirms that two original islands remain stable during this period, while other positions are experiencing different levels of settlement.

ABSTRACT

Detecting small-magnitude rates of vertical land movement (VLM) of the Earth’s surface using interferometric synthetic aperture radar (InSAR) is challenging, particularly when using a short time series of acquisitions.  An investigation into land subsidence in the Perth Basin (Western Australia) began in 2011 on the evidence that two continuously operating GPS stations (CGPS; Hillarys and Gnangara; IGS codes HIL1 and PERT) indicated land subsidence varying from 2 mm/yr to 6 mm/yr (Featherstone et al. 2012).  This appears to be due to changes in groundwater abstraction from the Perth Basin for domestic and agricultural use.

Insufficient historical geodetic observations of VLM (i.e., GPS, levelling and InSAR) were available to reliably determine the small-magnitude VLM over the Perth Basin, so a holistic geodetic monitoring program including repeat levelling, CGPS and InSAR was started (Filmer et al. 2013).  The InSAR component of the monitoring system comprises a German Space Centre (DLR) science project which began X-band SAR acquisitions from the TerraSAR-X (TSX) satellite mission at 11-day intervals over the Perth Basin (specifically the Perth metropolitan area) from October 2012.

Twelve TSX scenes were processed after the first six months of acquisition (April 2013).  Persistent scatterer (PS) processing using StaMPS (Hooper et al. 2004) and Doris (Kampes et al. 2003) for interferogram formation revealed many noisy PS, which masked reliable VLM detection at the sub-centimetre level (Garthwaite et al. 2013).  A second processing was undertaken after 12 months using 30 TSX scenes, but this time using the small baseline (SBAS) technique (e.g., Berardino et al. 2002).  We formed 95 SBAS interferograms, with the processed PS showing less noise, but line of sight (LOS) velocities reaching 10 mm/yr in magnitude.  These rates are larger than expected (acknowledging the geometric difference between VLM and LOS velocities) because CGPS-derived subsidence during this period was ~-2 mm/yr, so that the TSX LOS velocities are assumed to contain artefacts, primarily from atmospheric phase delay.

We now have 55 TSX scenes over two years (October 2012 - October 2014) which were re-processed using the SBAS technique as per the 12 month time series.  This large number of high-resolution scenes used a large amount of computer hard disk space (~5 TB) because almost the full spatial extent of each TSX scene was processed to cover as much of the Perth metropolitan area as possible.  To keep the required processing space to a minimum and also limit the length of the perpendicular and temporal baselines, the SBAS interferograms were selected manually.

Atmospheric phase delay still poses a problem, particularly over the land-sea boundary, although the longer time series facilitates the VLM PS phase becoming larger (and significant) relative to the atmospheric PS phase.  The large stack of SBAS interferograms also smooths the influence of the atmospheric phase in each PS.  Because 55 TSX scenes provide better redundancy than our earlier studies, we have investigated using an iterative outlier detection process.  This allows the identification and removal of scenes that contain the highest contamination from atmospheric phase delay from the interferogram stack.  The outlier detection process is expected to reduce the influence of atmospheric phase delay on the LOS velocities, lowering the uncertainty of the PS so that the derived VLM rates are more reliable.

REFERENCES

Berardino P, Fornaro G, Lanari R, Sansosti E (2002) A new algorithm for surface deformation monitoring based on small baseline differential SAR interferograms, IEEE Transactions on Geoscience and Remote Sensing, 40(11): 2375-2383

Featherstone WE, Filmer MS, Penna NT, Morgan LM, Schenk A (2012) Anthropogenic land subsidence in the Perth Basin: challenges for its detection and monitoring, Royal Society of Western Australia, 95(1):53-62

Filmer MS, Featherstone WE, Morgan L, Schenk A (2013) A future geodetic monitoring system for vertical land motion in the Perth Basin, Australia. Poster presented at EGU Symposia, April 2013, Vienna, Austria.

Garthwaite MC, Filmer MS, Schenk A, Lawrie S, Thankappan M, Dawson JH, Featherstone WE, Morgan L (2013) Space-geodetic determination of surface deformation due to resource extraction in Australia. In: Proceedings of TerraSAR-X science team meeting 10-14 June 2013, DLR – Oberpfaffenhofen, Germany. Online, http://sss.terrasar-x.dlr.de/papers_sci_meet_5/final/TSX_poster/23_TSX_science_meeting_2013_paper.pdf

Hooper A, Zebker H, Segall P, Kampes B (2004) A new method for measuring deformation on volcanoes and other natural terrains using InSAR persistent scatterers. Geophysical Research Letters, 31, Article No L23611. doi: 10.1029/2004GL021737

Kampes BM, Hanssen RF, Perski Z (2003) Radar interferometry with public domain tools. In Third International Workshop on ERS SAR Interferometry, `FRINGE03', Frascati, Italy, 1-5 Dec 2003, 6 pp

Mountainous areas cover about 64% of the Greek territory. High relief, heavy precipitation, seismicity, erosion and dense stream network are some of the conditioning and triggering factors of rock falls and landslide events in Greece.

The main objective of the present study is the detection and analysis of slow moving landslides, characterized as creep movements, pre or post failure landslide states by exploiting time-series surface deformation measurements at the Southern part of mountain range Pindus, the “backbone” of Greek mainland.

The area of interest covers 4000km² over two regional units, Evrytania and Aetolia-Acarnania, located in Central and in Western Greece respectively. Pindus zone extends in NW-SE direction. The geomorphological conditions of the region are intense, with sharp slope inclinations and the maximum elevation values often exceeding 2000 m as a result of the tectonic evolution, geological processes, climatic and hydrologic conditions. The heterogeneity of the various lithological units (limestones, chert, claystones and sandstones) as well as the intense tectonic deformation (close folds, nappes, dense fracturing) are conditions leading to slope instability.

The temporal coverage of the analysis extends between 1992-2010 by using descending tracks of ERS-1/2 and ENVISAT SAR data. We rely on the available high temporal sampling rate of the input satellite imagery in combination with strong filtering schemes, to overcome the inherent difficulties incurred from low coherence in vegetated mountainous regions. We employ Small Baseline Subset (SBAS) technique for the time-series analysis of the data, in order to extract ground velocities and associate them with slow-moving landslide events. The implementation of SBAS possesses numerous assets on the analysis of large scale surface and temporal evolution of deformations by overcoming a set of limitations. The lack of dominant point scatterers identified in the area of interest due to mountainous topography, slope inclination, aspect, vegetated areas and atmospheric contributions at high altitudes enhances the significance of SBAS technique as the most suitable tool for a spatially denser deformation mapping.

Interferometric velocities are subsequently correlated with ground truth records assessed in cooperation with the Institute of Geology and Mineral Exploration (The Greek Geological Survey). The integrated geo-information, namely velocity rate maps and time-series displacements with in-situ observations, reveal areas prone to landslides, their activity and intensity state, movement type, as well as their cause and frequency of occurrence. A fully updated landslide inventory map is developed for the region of interest.

At second phase, the complete aforementioned database constitutes the initiation point for quantitative analysis by means of data-driven and/or physically based methods towards spatial and temporal frequency determination of slow-moving landslides leading to susceptibility assessment and mapping.

This work was funded by the European Union 7th Framework Programme (FP7-REGPOT-2012-2013-1), in the framework of the project BEYOND “Building a Center of Excellence for EO-based Monitoring of Natural Disasters”, G. A. 316210.

Over the last 390 years, the City of Oslo shoreline has seen dramatic changes. Since the great fire of 1624, the city's waterfront has been pushed further and further out into the fjord. This development is culminating now with the Barcode Project, a dramatic set of five high-rise buildings constructed on former industrial and dock land in the central Oslo area known as Bjørvika.

Bjørvika surrounds the outlet of the river Akerselva. The area has been rising from the sea since the end of the last ice age due to isostatic uplift. The result is a thick set of marine and fluvial sediments that are still undergoing natural compaction. However, the most dramatic settlement is occurring within the anthropogenic fill material that has been used to expand the harbour area over the last centuries. This subsidence is having dramatic effects on infrastructure in the area, including the Oslo Central railway station. While most modern buildings are founded on piles that reach down to bedrock, many historic building, including the station, are built upon wooden foundations. Even if new buildings are stable, however, settlements in the surrounding area may affect infrastructure connections, such as water and power. Recently, it has been claimed that the Barcode development has accelerated subsidence in the area by altering the groundwater table and reducing the pore pressure in the sediments. A lowered groundwater table could also lead to degradation of wooden foundations that are now exposed to an oxidizing environment.

InSAR studies of the greater Oslo area using ERS data from the 1990's revealed the pattern of subsidence due to sediment/fill compaction as well as groundwater changes due to tunnelling activity (Dehls and Nordgulen, 2003a, 2003b, Lauknes et al., 2005). Recognizing the importance of this technique, we began to acquire Radarsat-1 data in both ascending and descending geometries in 2004. In 2008 we began using the Radarsat-2 satellite and have continued ever since, using standard mode. The Radarsat acquisitions have been possible by an agreement between the Norwegian Space Center and MDA to purchase data for Norwegian public authorities.

In November 2013, Airbus Defence and Space/Infoterra GmbH began tasking the TerraSAR-X and TanDEM-X satellites to acquire ultra-high resolution Staring SpotLight (ST) mode imagery over the Bjørvika area for technical evaluation. In this study, we present the results of PSI processing over the harbour and downtown core and compare with earlier ERS, Radarsat-1 and Radarsat-2 data spanning the last three decades. All data stacks have been processed using a Persistent Scatterer (PSI) algorithm implemented within the GSAR software package (Larsen et al., 2005), upgraded to support ST mode. A Lidar point cloud, supplied by the City of Oslo planning department is used to evaluate the accuracy of PS location calculation. Subsidence rates are compared to in situ measurements, including precision levelling campaigns that began in 1928.

Dehls, J. F. & Nordgulen, Ø., 2003. Analysis off InSAR data over Romeriksporten. NGU Report 2003.076.
Dehls, J. F. & Nordgulen, Ø., 2003. Evaluation of the use of PSInSAR for the monitoring of subsidence in Oslo region. NGU Report 2003.105.
Larsen Y, Engen G, Lauknes TR, Malnes E and Høgda KA (2006) A generic differential interferometric SAR processing system, with applications to land subsidence and snow-water equivalent retrieval. Fringe 2005 Workshop, Frascati, Italy.
Lauknes, T. R., Dehls, J., Larsen, Y., Høgda, K. A., & Weydahl, D. J., 2005. A Comparison of SBAS and PS ERS InSAR for Subsidence Monitoring in Oslo, Norway. Fringe 2005 Workshop, Frascati, Italy.

Western China is very susceptible to landslide hazards. As a result, landslide detection and early warning are of great importance in Western China considering the severe consequences that large-scale landslides may induce. The surface deformation measurement is often recognized as one of the most effective methods to monitor landslides and their movements. This work concentrates on the application of Persistent Scatterer Interferometry (PSI) Technique for detection and monitoring of large-scale landslides carried out in Li County, Sichuan Province, Western China. This work is part of an ongoing National Basic Research Program of China (973 Program), aiming to develop an InSAR approach to monitor the long time-series displacements and to identify slow moving landslides and potential areas where local authorities may assess the effectiveness of risk mitigation measures and quickly respond to possible landslide damage. In particular, a giant landslide, located in Li County of Sichuan, is a typical large-scale landslide and seriously affected by the “5.12” Wenchuan Earthquake in 2008. This landslide poses severe threats to 317 National Road and villages both on the slope and at the foot of the mountain. The Persistent Scatterer Interferometry is performed using both ascending and descending scenes acquired from TerraSAR-X SM mode since 2013 to get the spatial distribution of surface displacements of this giant landslide. All the datasets are processed in the mode PSInSAR and Small Baseline Subset (SBAS). The outputs of PSI help to redefine the landslide geometry, the state of activity and the spatial variance of velocities on the slope. The Multi-source spatial sensor network for in-situ monitoring of this landslide is also established on the slope, which helps to provide the deformation monitoring from diverse boreholes using different kinds of sensors such as GPS, inclinometer, Ground-based SAR and terrestrial laser scanning. In order to accomplish a more comprehensive validation result, ten corner reflectors (five for ascending orbit and five for descending one) are also installed near the boreholes. The outputs of PSI are integrated with CR-derived displacements and validated by in-situ sensors. In addition, the correlation analysis and accuracy assessment are performed. In particular, the displacements measured by inclinometers and GPS are projected to the direction of Line Of Sight (LOS) of the satellite, and shown to be consistent with the deformation acquired by PSI technique. Moreover, the integration of PSI displacements and ground-based monitoring data helps to retrieve a comprehensive and reliable surface displacement field, while a long time-series of displacements facilitates to better understand of landslide kinematics and to observe the relationship with triggering factors. The landslide ground surface displacements combined with detailed geological, geotechnical and geomorphological studies, allow a preliminary identification of the boundaries of areas with significant movements, thus providing reliable data support for the forecasting and monitoring of the landslide.

The Betic-Rif cordilleras are build-up in the western Mediterranean by the deformations related to the Eurasian-African plate boundary. The recent NW-SE convergence produces the simultaneous development of large folds and faults in the central part of the Internal Zones of the Betic Cordilleras, which continues active up to Present. The uplift of the mountain ranges is mainly related to the development of folds in this regional compressive setting. The most abundant faults recognised in surface along the central part of the Cordillera show normal slip, sometimes with dextral or sinistral components.

Sierra Tejeda constitutes a relief related to the development of a complex shaped recent antiform that is located near the coast line, overprinted on previous folded Alpujarride rocks belonging to the Internal Zones of the Betic Cordilleras. Along the northern limb of Sierra Tejeda is located the contact between the Internal and External zones. Geomorphologic evidences, like the important incision of the fluvial network or the cliff development along the coast line, indicate its recent uplift. The southern limb of the antiform is deformed by a set of WNW-ESE oriented faults dipping towards the SW and with a normal and transcurrent regime.

At the northern limb of Sierra Tejeda is located the Zafarraya fault, a high-angle northwards dipping normal fault, which produces the development of a half graben in its hanging wall filled by Miocene-to-Quaternary sediments. In the central part of the Betic Cordillera, the Zafarraya fault constitutes one of the structures with evidence of very recent tectonic activity. There are reports of displacements along this fault during the 1884 Christmas Andalusia earthquake (felt intensities of IX-X and estimated magnitude between 6.5 and 7) that destroyed part of the villages of the region. It is the most recent destructive earthquake of the Iberian Peninsula that caused about 800 fatalities. Field studies determine that the fault has a total length of more than 15 km. Its trend varies along the strike, approximately E-W to the south of Zafarraya and curves to a NW-SE orientation at its western end, where deformed rocks of the External Zones are found. The fault plane in the Zafarraya region dips 60º northwards, and its striae indicate that it is a normal fault. Fault slip is of several hundred of meters.

In the central Betic cordilleras, folds and faults developed together since Middle Miocene. Reverse blind faults —probably located at depth— and large folds are responsible for the relief uplift, and are directly associated with the oblique convergent character of the plate boundary. However, the active faults that are recognised in the upper part of the crust in this transect of the Cordilleras are mainly normal and have large vertical displacements.

This paper investigates the applicability of satellite radar interferometry using ERS SAR and Envisat ASAR datasets to help to constraint the geological hazard in this part of the Betic Cordillera.

The East African Rift is an ~6400 km long, 50-60 km wide plate boundary formed by the divergence of the Nubian and Somalian plates. Archive Envisat data has been used to study deformation associated with a range of plate boundary processes: including dyke intrusions from Afar to Tanzania (e.g. Wright et al., 2006, Calais et al., 2008), earthquakes throughout the rift (e.g. Baer, et al., 2008) and deformation at 12 volcanic centers (e.g. Biggs et al., 2011). However, there are still many unanswered questions, for example, how is strain partitioned between magmatic and tectonic processes? What drives deformation at non-erupting volcanoes? Furthermore, only a handful of the identified volcanoes are routinely monitored, and so satellite data has an important role to play in volcano monitoring.

The launch of the new, dedicated InSAR satellite, Sentinel 1a, will allow a much more thorough investigation into the mechanisms which facilitate this rifting and control volcanic deformation. Here I will present preliminary results from early Sentinel satellite data, focusing on deforming volcanic centers, and any seismic events.

 

Baer, G. Y. Hamiel, G. Shamier, R. Nof., 2008. Evolution of a magma-driven earthquake swarm and triggering of the nearby Oldoinyo Lengai eruption, as resolved by InSAR, ground observations and elastic modeling, East African Rift, 2007. Earth and Planetary Science Letters, 272(1-2), pp.339–352.

Biggs, J. I. D. Bastow, D. Keir, and E. Lewi (2011), Pulses of deformation reveal frequently recurring shallow magmatic activity beneath the Main Ethiopian Rift, Geochem. Geophys. Geosyst., 12(9).

Calais, E. N. d'Oreye, J. Albaric, A. Deschamps, D. Delvaux, J. Déverchère, C. Ebinger, R. W. Ferdinand, F. Kervyn, A. Macheyeki, A. Oyen, J. Perrot, E. Saria, B. Smets, S. Stamps, C. Wauthier., 2008. Strain accommodation by slow slip and dyking in a youthful continental rift, East Africa. Nature, 456(7223), pp.783–7.

Wright, T.J. C. Ebinger, J. Biggs, A. Ayele, G. Yirgu, D. Kier, A. Stork., 2006. Magma-maintained rift segmentation at continental rupture in the 2005 Afar dyking episode. Nature, 442(7100), pp.291–4.

Located at the northern part of Taiwan, Taipei is the most densely populated city and the center of politic, economic, and culture of this island. North of the Taipei basin, the active Tatun volcano group with the eruptive potential to devastate the entire Taipei is only 15 km away from the capital Taipei. Furthermore, the active Shanchiao fault located in the western margin of Taipei basin. Therefore, it is not only an interesting scientific topic but also a strong social impact to better understand the assessment and mitigation of geological hazard in the metropolitan Taipei city. In this study, we use 12 high resolution X-band SAR images from the new generation COSMO-SkyMed (CSK) constellation for associating with leveling and GPS data to monitor surface deformation around the Shanchiao fault and the Tatun volcano group. The stripmap mode of CSK SAR images provides spatial resolution of 3 m x 3 m, which is one order of magnitude better than the previous available satellite SAR data. The higher resolution leads to an increase of the density of the measurable targets relative to those retrieved from medium resolution datasets (C- and L-band). Besides, the sensitivity of displacements is increased with shorter wavelengths, which enhance the capability of detecting very slow displacements rates. Furthermore, the more frequent revisit of the same Area of Interest (AOI) of the present X-band missions provides massive datasets to avoid the baseline limitation and temporal decorrelation to improve the temporal resolution of deformation in time series. After transferring the GPS vectors and leveling data to the LOS direction by referring to continuous GPS station BANC, the R square between PS velocities and GPS velocities is 0.96, which indicates the high reliability of our PSInSAR result. In addition, the well-fitting profiles between leveling data and PSInSAR result along two leveling routes both demonstrate that the significant deformation gradient mainly occurs along the Shanchiao fault. The severe land subsidence area is located in the western part of Taipei basin just next to the Shanchiao fault with a maximum of SRD rate of 30 mm/yr. However, the severe subsidence area, Wuku, is also one industrial area in Taipei which could be attributed to anthropogenic effect. In the future, we will use all available images to monitor the temporal and spatial variation in deformation to better understand the activity of the Shanchiao fault.

Multi-temporal SAR Interferometry (MTInSAR) techniques allow detecting and monitoring millimetric displacements occurring on selected point targets exhibiting coherent radar backscattering properties. Successful applications to different geophysical phenomena have been already demonstrated in literature.

During the last several years new application opportunities have emerged thanks to the greater data availability offered by recent launches of radar satellites, and the improved capabilities of the new space radar sensors in terms of both resolution and revisit time. Currently, different space-borne SAR data in L-,C- and X-band are available for InSAR applications.

The archived data from the European Space Agency (ESA) missions ERS-1/2 and ENVISAT (ENV) acquired in C-band (about 5 cm of wavelength), at medium resolution (5x20 m2) from 1992 to 2011 make available a large number (> 40) of images covering more then 10 years with a minimum revisit time of 35 days, which allows performing ground instability analysis back in time almost all over the Earth. ESA Sentilnel-1 constellation will provide the continuity of the C-band SAR operational applications from October 2014.

Thanks to the technological maturity as well as to the wide availability of SAR data, MTInSAR can be used to support systems devoted to environmental monitoring and risk management. 

This work present results obtained in the framework of two projects:  the CAR-SLIDE (Mapping and monitoring system for landslides forecast) project which is funded by MIUR (PON R&C 2007-2013) and ADF (Archiving Data Fusion) project, funded by the Italian Space Agency (Second ASI Call dedicated to SMEs).

CAR-SLIDE  is aimed at implementing an advanced diagnostic system capable to warn of and monitor landslide events along railway networks, by integrating in situ data detected from on board sophisticated innovative measuring systems, with Earth Observation (EO) techniques. In particular SAR interferometry is used to monitor landslide events. The CAR-SLIDE system is structured in a modular way and mainly consists of the following three subsystems (S/S): (1) railway infrastructure monitoring S/S (based on video inspection sensors, georadars, vibration measurement systems and track geometry measuring devices) ; (2) Environmental Monitoring system (based on satellite SAR and optical data, and weather forecast provided by prognostic non-hydrostatic numerical mesoscale models); (3) Support Decision System (SSD). All measurements provided by the respective monitoring systems are integrated into the SSD, with the aim of improving the railway transport security management, by extending the concept of “black-box for alarm generation” to the actual idea of “telemetry for security management”.

The goal of the ADF project is to design an innovative system for EO data acquisition, processing, fusion and archiving to support public institutions and private companies involved in emergency management, territorial planning and precise farming.  The system has the capability to interface with many EO (SAR and optical) and traditional data sources and archives, thus enabling their selection and processing on demand, and to perform data fusion of EO data with ancillary data (INSPIRE, OpenStreetMap, etc.), with a high level of automatism. Concerning emergency management, data fusion techniques of SAR and optical data, together with land cover maps and digital terrain models are used to provide a priori estimations of Persistent Scatterers density along different line-of-sights, look sides  and pass directions.

We presents results obtained by processing COSMO-SkyMed satellite data acquired over Calabria region (Southern Italy)  and Liguria region (Northern Italy).

The first test case shows interesting results concerning the Calabria's Tyrrhenian railway line, classified as fundamental line by the Italian Rail Network and belonging to  the Trans-European Transport Network. Moreover, it provides another interesting example of infrastructure monitoring concerning the “Giambarelli” viaduct along the A3 highway close to the Barritteri town.  

The second example consists in a retrospective analysis on the detection of precursory signals related to the landslide which occurred on January 2014 close to the town of Marina di Andora. The landslide caused the derailment of a train and the interruption of the railway line connecting north-western Italy to France. We show that a cluster of moving targets coincides with the structures (buildings and terraces) affected by the 2014 landslide. The analysis of the MTI time series further shows that the targets had been moving since 2009, and thus could have provided a forewarning signal about ongoing slope or engineering structure instability.

Further examples, concerning the suitability of MTInSAR techniques and high-resolution SAR sensors for man-made structures stability over the Genoa metropolitan and surrounding areas, are also presented and discussed.

 

Acknowledgments

Work supported by the projects “Landslide Monitoring and Mapping System - CAR-SLIDE” (PON 01 00536) and “ADF - Archiving Data Fusion” (Second ASI Call dedicated to SMEs). CSK data are provided by ASI (Agenzia Spaziale Italiana) in the framework of CAR-SLIDE and ADF projects and the collaboration between Geophysical Applications Processing - GAP srl and CIDOT (Centro di eccellenza di Interpretazione Dati di Osservazione della Terra – ASI Matera).

Sited in the Northern part of the Apulia Region, between the Lesina Lagoon and the mouth of the Fortore River, Lesina Marina lies not far from the Pietre Nere (Black Stones) Point, a very interesting site and the  only outcrop of magmatic rocks in the Puglia region and in the whole Adriatic coast of Italy.

The excavation of a canal through this area, performed in 1930, exposed grey micro- and meso-crystalline gypsum with intercalations of black limestones and marls of Upper Triassic age, mantled by loose sandy Quaternary deposits. The gypsum bedrock shows a high density of cavities, either dissolutional conduits or voids related to gravitational collapse processes [1].

A large dataset, worth 20 years of data from multiple sensors (ERS, ENVISAT, Radarsat-2,COSMO-SkyMed), covering a total time interval from 1992 to 2012, with varying ground resolutions and repeat times, was processed through Persistent Scatterers Interferometric Synthetic Aperture Radar (PSInSAR) techniques. For legacy sensors (ERS and ENVISAT), ascending/descending combinations allowed separating East-West and Up-Down components of displacement [2]. Data were cross-compared through a multi-resolution analysis in order to assess the displacement rates.  Results show the presence of steadily uplifting PS objects over the Lesina Marina urban area, with maximum values located around the Acquarotta canal, exhibiting average uplift velocities reaching 4-5 mm/year, and decreasing smoothly towards the W-SW to about zero, with locally slightly negative values. Overall velocity offset values are set based on statistical grounds, by comparing results from full-frame PSI results processed through slightly different algorithms.

High-precision leveling measurements, performed on several tens of topographic points on the study area, during two ground truth campaigns, in 2000 and in 2010, are compatible the presence of upward movements, although the leveling stations are located mainly close to the canal, in a strip roughly oriented in a N-S direction, about 300 m wide.

The area affected by the vertical displacements coincides with the higher part of the salt dome, which was pushed out by several magmatic dikes (Punta delle Pietre Nere formation), 60 My aged. Therefore a residual diapirism cannot be excluded [3]. Another phenomenon is the hydration of the residual anhydrite, in the core of a gypsum mass residing beneath or close to the surface. Gypsum hydration, which involves a volume increase of up to 63% of the hydrating mass [4], has been observed to cause consistent and sudden uplift phenomena on other sites, such as the city of Staufen, Germany [5], causing serious infrastructure damages. In the present case, the uplift phenomenon is much less intense, but the uplift intensity is apparently increasing towards the canal; this could be congruent with a smaller mass of anhydride, not directly exposed, or with a more ancient process, both hypotheses leading to a slower hydration process rate. Rock analysis data collected from both borehole and surface sampling sites reveal the presence of anhydrite, which would therefore confirm the hypothesis of uplift due to bulging caused by halite hydration.

We also find a small number of detected PS close to a probable sinkhole site, showing relative downward motion, compatible with displacements measured locally by nearby leveling benchmarks, suggesting the possibility of precursory movements which could be associated to sinkhole collapse phenomena.

Acknowledgements

Work supported by the RITMARE CNR flagship project, and from the Italian Ministry of Research in the framework of the PRIN 2008 research grant. ENVISAT and ERS data provided by European Space Agency Category-1 project N. 5367, “Subsidence monitoring in Daunia and Capitanata (Puglia Region, Italy) through multi-temporal point-target DInSAR techniques”. RADARSAT-2 Data and Products @ MacDonald, Dettwiler and Associates Ltd. 2014 – All Rights Reserved. RADARSAT is an official trademark of the Canadian Space Agency. COSMO-SkyMed data are provided by ASI (Agenzia Spaziale Italiana) in the framework of the RITMARE research project.

References

[1]  Fidelibus, M. D., Gutiérrez, F., Spilotro, G, “Human-induced hydrogeological changes and sinkholes in the coastal gypsum karst of lesina marina area (Foggia province, Italy)”. Engineering Geology 118, 1–19, 2011.
[2]  Refice A., Bovenga F., Pasquariello G., Fidelibus D., Spilotro G., “Bulging of the gypsum dome at Lesina Marina (Southern Italy) revealed by PSI techniques”, Proc. FRINGE 2011, ESA SP, vol. SP-697, ISBN: 978-92-9092-261-2, ISSN: 1609-042X, ESA-ESRIN, Frascati (RM), Sept. 2011.
[3]  Mastronuzzi G., Sansò, P, “Holocene uplift rates and historical rapid sea-level changes at the Gargano promontory, Italy”, Journal of Quaternary Science 17 (5-6), 593–606, 2002
[4]  Bell, F., Cripps, J., Culshaw, M, “A review of the engineering behaviour of soils and rocks with respect to groundwater”, Engineering Geology Special Publications 3, 1–23, Geological Society, London, 1986
[5]  Motagh, M., Wetzel, H.-U., Breuer, B., Anderssohn, J., Sass, I., Kaufmann, H., 19-23 September, 2011. Radar interferometry time series analysis to evaluate the source modeling of the city uplift in staufen i. br. (germany). Proc. FRINGE 2011, ESA-ESRIN, ESA SP, vol. SP-697, ISBN: 978-92-9092-261-2, ISSN: 1609-042X, Frascati, Sept. 2011.

Mud volcanism is a process generally connected to hydrocarbon traps and consists in the extrusion of material mobilized from the subsurface, such as gases, saline waters, mud and fragments of country rocks. Mud volcanoes are conical edifices of various sizes (from several centimeters to a few hundred meters in height and several kilometers across), with morphology similar to those of magmatic volcanoes. In this work, the Differential SAR Interferometry technique is used to investigate the ground deformation related to the activity of the mud volcanoes of Azerbaijan during the period spanning from October 2003 to November 2005. This study is based on the elaboration of 9 ENVISAT images and on the analysis of a set of wrapped and unwrapped interferograms, selected according to the best coherence values, which allowed the detection of significant ground displacement due to the activity of several mud volcanoes (Antonielli et al., 2014; Tectonophysics, in press). In particular, this work focuses on two important deformation pre-eruptive events at Ayaz-Akhtarma and Khara-Zira edifices. In general, the ground deformation at mud volcanoes is originated by pressure and volume variations in the fluid reservoir, caused by the inflation or deflation of the mud-fluid mix. In our study cases, the deformation pattern is characterized by pre-eruptive inflation that reaches a cumulative value of up to 20 cm at the Ayaz-Akhtarma, and up to 10 cm at the Khara-Zira mud volcanoes in about two years. Similar bulging before an eruption has also been observed at magmatic volcanoes, where uplift is typically associated with magma intrusion and/or fluid injection in the shallow crust. It is worth noting that a sector of Ayaz-Akhtarma edifice is affected by subsidence, which might be related to a deflation zones coexisting with the major inflation zone. Also in some magmatic calderas volcanic uplift may be concurrent with the presence of areas of subsidence and may generate complex ground deformation patterns. Satellite radar interferometry has proved to be suitable for the study of mud volcanoes activity and shows that mud and magmatic volcanoes display some similarities in ground deformation behaviors.

The Kivu rift Basin is located in the western branch of the East African rift. The region is characterized by intense seismicity, active volcanism, and in many areas, slope instabilities are a constant concern even increased by climatic changes. The region is also densely populated and people are undergoing the severe consequences of permanent geopolitical unrest. Internal migrations towards urban centers, refugees fluxes during conflicts and settlements in peripheral areas regularly increase the pressure on land. In certain areas were the natural risk raises sometime well above fatal thresholds, catastrophic events such as major landslides often come on top the news. 

1. The Nyiragongo eruption as a research triggering

In 2002, the Nyiragongo volcano was responsible of an important disaster: known for hosting the largest active lava lake on Earth, a fissure open on its southern flank and drained the lava lake into Goma, ravaging a significant part of the city in a few hours. On top of the direct consequences of the lava flows, the long-term consequences of that catastrophic event are still felt today as they deeply and durably impacted the socio-economical live of the region.

At the earliest stage of the disaster, an international mobilization setup followed by successive initiatives focused on the support of the local authority in charge of the volcano monitoring. Besides these structural supports, scientific research also intensified and focused on the study and understanding of the processes involved. 

2. Ground-based versus remote sensing

Important actions were essentially achieved in the field: reinforcing the seismic monitoring, implementing other networks such as tiltmeters, EDM, gas and temperature, monitoring the fissure opening, etc. But the difficult situation rapidly discouraged the international teams who were also involved at various levels in local scientific training.

In 2005, a Belgium-Luxemburgish collaboration started, specifically dedicated on the study of the feasibility of InSAR on tropical African active volcanoes. In the framework of that project named “SAMAAV” (see d’Oreye et al, FRINGE 2007), a regional ground deformation associated to the Nyiragongo eruption was detected; it was the beginning of a systematic monitoring of the volcano using various SAR sensors. Other sensors were also used e.g. for the monitoring of the SO2 volcanic plume.

3. Premises of integrated approach

That first exclusive remote sensing approach not only motivated further RS projects such as “Vi-X” which aimed at creating high resolution DEM using TanDEM-X and TERRASAR-X data and develop the Splitband interferometry technique (see Derauw et al, this meeting). It also motivated the setup of a first integrated project: “GORISK” aimed at intensify the study of ground deformations using InSAR, GPS, tilt, and also address passive degassing and plume impact on health issues (van Overbeke et al, 2010).

Results confirmed the relevance of using InSAR: Wauthier et al (2012) proposed the best-fit model for the deformation associated to the eruption of 2002. The model suggested a combination of two dykes: the surficial one responsible for the lava flows, and a deeper one extending far southwards under the Lake Kivu and raising new concerns about possible interactions with that dissolved gas layered lake.

4. From single- to multi-hazards

Through important commitment in the field, “GORISK” evidenced also the strong need in considering the volcanic hazard from various angles. On the one hand, it evidenced the gaps in monitoring the seismicity regionally, either because no regular collaboration existed between Rwanda and DR Congo, or simply because no regional network were operational. On the other hand, it appeared that risk management imposed to address other main hazards affecting the region as well as vulnerability issues.

“GeoRisCA” project was designed to assess the main hazards affecting the region: volcanoes, landslides and seismicity, and to assess the vulnerability of population with regards to these hazards.

From the local scale of the volcanic affected site, the interest grew to regional scale, from the Virunga active volcanoes to the North of the Tanganyika Basin were seismicity and slope instabilities are affecting urban areas of Bujumbura (Burundi) and Bukavu (DRC).

5. Studying the variations

Recently, correlations were discovered between InSAR, GPS, seismic and volcanic plume measurements weeks before the 2010 Nyiragongo’ neighbour, the Nyamulagira (Smets et al, 2014). That has motivated the onset of a new “RESIST” project, which concentrates on the study of variation of monitored parameters and involves the new generation of SAR satellites. “RESIST” relies strongly on the reinforcement of the seismic monitoring capacities, and on intensive SAR imaging. SENTINEL will play a major role for the study of ground deformations associated to the volcano-tectonic activity, but it will also be involved in the detection of slope instabilities over the whole area using the wide swath mode of SENTINEL.

6. Conclusions

The experience acquired evidences the close link that exists between science and society in that specific and difficult context of Central Africa. This communication summarizes the frame and main accomplishments of each initiative, shed light of the current state of the art, and underlines how remote sensing, field based techniques, local training and capacity building are related to turn all these scientific research efforts into sustainable steps towards development.

The city of Lisbon is one of the most important cities in Portugal. Besides being the capital of the country, it is also an important governmental, business and historical place, with 3 million people living in the city and its surroundings. Lisbon is located in a region of high seismicity due to its proximity to the joint point of the Eurasian and African tectonic plates, having been affected by devastating earthquakes and tsunamis in the past. It is known that some areas of the city present an instable behaviour due to their geomorphologic properties such as the existence of alluvial sediments transported by underground streams and the constant presence of groundwater. Besides, as the city has been expanding in the last decades, an increase in the terrain load due to the buildings and infrastructure construction has been verified, aggravated by the construction of subway lines and groundwater extraction.

This study presents an application of PS-InSAR to the city of Lisbon in order to perform a regional analysis of soil instability. An analysis of non-linear displacements over a dataset of Envisat images using the SARPROZ software is proposed. This method confirmed the existence of subsiding areas identified in previous studies (Heleno et al., 2011), but also enabled the detection of new unstable areas. Several areas presenting instability were identified in the city, including the Downtown area, where an oscillating behaviour consistent to aquifer recharge during winter or to the infiltration of river water related to tides was verified. Subsidence behaviour reaching velocities above 18 mm/yr along the sensor line-of-sight was detected in landfill areas close to the river and even the railway line connecting Lisbon to Cascais presents signs of instability. Besides the regional analysis, more localized processing was performed for selected structures, such as MEO Arena building, in which a seasonal behaviour was identified. The results obtained through PS-InSAR were compared to data from geodetic measures.

 

Reference

Heleno, S.I.N., L.G.S. Oliveira, M.J. Henriques, A.P. Falcão, J.N.P. Lima, G. Cooksley, A. Ferretti, A.M. Fonseca, J.P. Lobo-Ferreira, and J.F.B.D. Fonseca (2011). “Persistent Scatterers Interferometry detects and measures ground subsidence in Lisbon”, Remote Sensing of Environment, 115, pp. 2152-2167.

An interface is presented which allows setting up and lauching inversions of InSAR data in order to determine the geometry, location and stress changes of sources such as magma intrusions, magma reservoirs, and faults beneath realistic topographies. The inversions [Fukushima et al., 2005] combine 3D elastostatic boundary element models [Cayol and Cornet] and a neighborhood optimization algorithm [Sambridge, 1999a, 1999b] to simultaneously determine non-linear model parameters (source geometry and location) and linear model parameters (source stress drop) and assess marginal probability density functions. Previously, the steps leading to setting up an inversion were tedious and time consuming. The interface allows saving time by guiding the user through the various imbricated steps required for an inversion. These steps are the definition of a model, the design of a topography mesh, the data subsampling, the covariance matrix creation, definition of the near neighborhood algorithm parameters and results visualization. Imbrications concern several aspect of the model definition: for instance, the type of source (fracture or reservoir) conditions the density and dimension of the topographic mesh; when eruptive dikes are modelled, surface fractures are needed for both the dikes and the topographic mesh. The interface also incorporates several rules of thumb needed for an accurate modelling. A graphical tool allows definitions of surface fractures from the interferograms. The model and data subsampling are graphically controlled at each step. Use of the graphical interface reduces the time needed to set up an inversion by a factor of 20. Inversions can then be launched from the interface on a cluster of PC operating under linux, further reducing computation time. Application of the interface to the Piton de la Fournaise October 2010 eruption is finally shown. This eruption was captured by two TerraSAR-X interferograms. We find that the displacement can be explained by a curved intrusion, which starts as a sill beneath the summit and becomes vertical at the location of the eruptive fissure.

From july 2006 to february 2007, three summit eruptions of the Piton de la Fournaise (Réunion Island) occurred. On march 30^th2007, a dyke opening triggered an eruption outside the central cone. On april 2^d, another eruption of exceptional intensity started 7 km southeast of the central cone and lasted until may 1^st. This eruption was associated to a collapse of a summit caldera by more than 300 meters and by the largest emitted volume of the XX^th and XXI^th century at the volcano.

Important eastward and downward displacement of the eastern flank of the volcano were recorded by InSAR data from ALOS and ENVISAT satellites during the april 2007 eruption. Co-eruptive interferograms show up to 1.4 meters of displacement towards the east, and a general subsidence up to 45 centimeters interrupted by around 35 centimeters of inflation in the central part of the eastern flank. This movement continued during more than one year after the end of the eruption, with up to 19 centimeters towards the east and 34 centimeters of subsidence from may 2007 to september 2008. In this work, we try to understand which structure caused the eastern flank displacement during and after the april 2007 eruption by doing inversion of the InSAR data. Our purpose is twofold : assess the hazard associated to flank displacementand understand the origin of the distal location of the eruption. We focus on the post-eruptive data as they include neither the march 30^th dyke displacement, nor the summit caldera collapse, making the pattern of deformation of the eastern flank less complicated.

C-band ENVISAT interferograms from six different lines of sight (3 ascending and 3 descending) were temporally interpolated to cover a common period from may 2007 to july 2008. A mixed Boundary Element Method combined to a Neighborhood Algorithm is used to invert the surface displacement. We make the hypothesis that the volcano is homogeneous and obeys to the linear elasticity theory. The real topography is taken into account using a Digital Elevation Model. The fracture geometry is supposed to be a curved quadrangle discretized in triangular elements and defined by 18 non-linear parameters. Linear parameters such as pressure and shear stress, with an unknown pitch are also inverted. Both stress changes are assumed homogeneous in a circular zone of the fracture plane. The Neighborhood Algorithm is an efficient Monte Carlo method to explore muti-parameters spaces and analyse misfit functions which have multiple minima.

Inversion results show that a very shallow fracture reproduces well the first order displacement of the eastern flank. The fracture is trapezoïdal with its longer side at low altitude and a low dip angle to the East. It roughly links the summit cone to the eruptive fissure of april 2007 which corresponds to the expected magma path. The fracture is submitted to an eastward slip up to around 30 centimeters. Second order displacement are not explained by this method and correpond to residuals of a few centimeters. This may be due to the hypothesis of homogeneity of pressure and shear stress on a circular zone of the fracture. To test this hypothesis, we further compare this result with a linear inversion of slip distribution on a fracture of the same geometry, assuming an elastic half space.

Landslides rarely behave as a rigid body that responds instantly to stresses. More frequently, they are characterised by a strain pattern (hence, displacements), increasing over time. Therefore, time-dependent failure relationships must be considered to describe the long-term deformation pattern known as “slope creep” which represents the slow deformation of slopes involving soil and rock, occurring under gravity and external loading. In the creep model, the materials show a particular evolution of the strain pattern characterised by three phases: a quasi-sudden deformation (primary creep), a steady-state deformation (secondary creep) and finally, a tertiary creep, in which continuous strain acceleration leads the material to failure.to failure.

Following the theory of the observational approach, several authors found that displacement is the best parameter to get information about the time of failure (ToF). In this perspective, many efforts have been carried out to attempt to predict landslide failure using semi-empirical models based on the slope creep theory. The continuous collection of displacement data is a key requirement for the application of semi-empirical models. Therefore, displacement-monitoring systems play a relevant role in landslide failure prediction. In the observational approach, the availability of displacement data characterised by a good distribution in space (i.e., the capability of attaining information about several portions of the slope under investigation) and high sampling frequency in time is a crucial point. Consequently, the role of monitoring is essential in collecting accurate and detailed data in both space and time and operational features of the monitoring systems are a key point. In particular, the temporal resolution of data is very important for properly detecting the acceleration stage for a slope suffering creep evolution over time. In this regard considerable progress has been made in recent years both with contact and remote sensing apparatus.

In order to consider the efficacy of Satellite A-DInSAR (Advanced Differential SAR Interferometry) methodologies for forecasting purposes, two features have been considered:

i) the ability to observe non linear deformation trend over time (i.e., acceleration, typical of the final stages of creep when the transition from secondary to tertiary creep leads to acceleration of displacements);

ii) the ability to "follow" the evolution of the observed landslide over time to properly describe the process without losing information.

The first one is mainly dependent on the sensor characteristics and on the data processing procedures, while the second one depends on the satellite revisit time.

Based on the displacement time series until failure of more than 40 historical landslides, inferred from the scientific literature, a theoretical analysis about the potential for the ToF prediction has been performed. The key parameter used in the analysis has been the "Revisit Time Factor" (RTF), given by the ratio between the Tertiary Creep duration (TC) and the Revisit Time (RT). More specifically, it has been assumed that the accelerating behaviour of the landslide can be identified and monitored only in case that two images at least have been acquired during the TC period, but the time of failure is “predictable” only starting from three images acquired within the TC. Based on this assumption the RTF has been computed for each landslide available by considering the nominal revisit time of past, present and future SAR satellites sensors (i.e. ERS 1/2, Envisat, TerraSAR-X, ALOS Palsar, ALOS Palsar-2, COSMO-SkyMed, Radarsat, Sentinel-1 and future constellations), thus achieving the A-DInSAR forecast potential for each landslide.

As expected, present and future satellites provide best scenarios in terms of theoretical predictability. The significant improvement in terms of revisit time, provided by present and future satellites, in fact, greatly increases the capability to monitor such processes. In this regard, it should be observed that 13 landslides could be predicted by past satellites characterized by long RT (namely, ERS 1/2, Envisat, ALOS PALSAR). A higher number of landslide (about 30) would be predicted by satellites with a shorter RT such as TerraSAR-X, COSMO-SkyMed constellation and Sentinel-1. However, the best results are only possible using the lowest possible revisit time of present and future constellations (COSMO-SkyMed, Radarsat Constellation and Sentinel-1). Under these conditions, though theoretical, almost all landslides would be predictable.

The obtained results allow in fixing some points concerning the application of SAR data for monitoring purposes by defining the theoretical, practical and functional limitations in direct relation to the dynamics of landslides. Finally, this approach integrates well with other works reported in the literature and aimed at defining the application of A-DInSAR methods basing on topographic and land use data in areas subjected to landslide issues.

The aim of this exploratory research is to capture spatial evolution and ground displacement trends caused by natural and anthropogenic factors in the Bucharest metropolitan area using sets of single polarized synthetic aperture radar (SAR) satellite data and multi temporal radar interferometry (InSAR) methodologies.  Three sets of SAR data acquired during the years 1992–2013 from ERS-1/-2 and ENVISAT, and recent TerraSAR-X satellites were used to provide maps of line-of-sight displacements.  This research considered two hypotheses.  The first hypothesis focuses on identifying possible trends in developments over former marshes, clay pits and landfill excavation areas, especially those located in the urban areas that displayed a chaotic development during the 1990-2000 period.  The second hypothesis considers increases in deep groundwater pressures after 2000, when important water consuming industries, such as industrial park areas, were shut down.  The analysis of the groundwater effects focused on changes that occurred before and after 2000, by looking at changes in groundwater elevations in 21 water wells, of which 13 are at industrial sites.  Seventy industrial areas were analyzed from which 14 represent large industrial parks.  These industrial parks were divided into industrial platforms that contain water wells and industrial platforms that do not contain water wells.  The satellite-based remote sensing results were integrated with results derived from classical methodologies (i.e., diachronic cartography) and field research.  The diachronic analysis was based on detailed historical maps from 1852 to the present (i.e., Borroczyn map, 1852; Szatmary map, 1864; the Bucharest City Plan, 1911, 1921, 1940; topographical map, 1980), orthophotomaps (2006, 2008, 2010) and a 2008 high resolution LiDAR digital elevation model (DEM).  The cartographic material facilitated the identification of highly dynamic urban areas which were then analysed for possible displacement trends.

In order to test the influence of the weather component in the displacement trends, we performed a statistical analysis of displacements in relation to daily air temperature and precipitation for the entire period covered by TSX images.  By implementing the Granger causality test, we obtained one-directional causality (of type Tmax and/or Tmin variables—Granger cause—displacement) for all displacement series. Ninety-two percent of the variation of displacements can be explained by assuming a spline smoothing nonlinear model and only 46% by assuming a linear regression model (53% by assuming a quadratic model).  The residuals, resulting from the best fitted model to the data, namely, spline and cubic-spline nonlinear regression, were not independent and contained important information.  The results suggest that the topographical variation has causes other than weather and should be better explained by more complex nonlinear variables.  Resulting displacement maps showed the possible relationships with (a) historical areas that experienced changes between 1800 and 1950 (i.e., developments over former marshes, clay pits, and landfill excavation areas), (b) changes that occurred before and after 2000 during the city’s most rapid development of build-ups and several industrial park areas.  The ongoing research efforts will focus on result validation using traditional geological, geomorphological investigations, and comparison with field GPS data (in case of the TSX-based displacement map product).

An intense earthquake swarm of more than 30000 events took place in Harrat Lunayyir (western Saudi Arabia) between April and July 2009. The swarm produced 19 earthquakes larger than magnitude 4, with the largest event of size 5.7 occurring on 19 May. Earlier studies have shown that the extensive ground deformation that followed the earthquake activity was caused by a magmatic dike intrusion that almost made it to the surface. These studies did not, however, report fully the spatial and temporal evolution of ground deformation during majority of the activity. Here we use additional Envisat data from both image and wide-swath modes to unravel the sequence of events during the intrusion and to analyze the interaction between the intruding dike and normal faulting at Harrat Lunayyir. We find that surface deformation is first detected in data from 16 May, about four weeks after the earthquake activity started, when surface displacements due to a ~2 m thick dike intrusion are observed. From 19 May, the dike propagated to within ~1 km of the surface, with graben-bounding normal fault displacements dominating the near-field deformation. The volume of the dike increased gradually from a total of ~0.07 km³ on 16 May to ~0.13 km³ on 17 June. Coulomb failure stress-change calculations of the dike-fault interaction show that the growing dike caused extension above its tip and triggered slip on the graben-bounding normal faults. In turn, the faulting lead to compression around the upper tip of dike, holding back its vertical propagation. Our observations suggest that magma is transported effectively by dikes from depths below the crust to shallow depths and the surface in the off-rift volcanic provinces in western Saudi Arabia, without significant resident time in the crust.

The main goal of this work is an analysis of vertical displacements (subsidence and uplift) on highly urbanized industrial areas using two satellite radar interferometry methods: Differential Interferometry Synthetic Aperture Radar (DInSAR) and Persistent Scatterer Interferometric Synthetic Aperture Radar (PSInSAR). To detect ground deformations using the DInSAR method a pair of satellite images is needed, while using PSInSAR method many satellite images (even tens) are needed. In this work a scope of applicability of both methods is described and detected vertical displacements are compared, which do not overlap with each other. Only the combination of results allows to present a complex map of vertical displacements in the study area. Satellite images in this work cover the Upper Silesian Coal Basin (USCB) area located in southern Poland. Images are provided inter alia by European Space Agency and were recorded between 1990 and 2001. Study area is highly urbanized, because of the presence of many active stone coal mines. Vertical displacements presented in work are conducted by underground mining exploitation and neotectonic movements. In the main, the DInSAR method allows to detect big (a few or several centimeters) and intensive (formed in a few days, 35-140 days) vertical displacements conducted by an underground mining exploitation. In the areas of the detected vertical displacements by using the DInSAR method the coherence is high, due to fact that those areas are highly urbanized (many permanent scatters occur).  The PSInSAR method in the main allows to detect small and permanent vertical displacements, which are correlated  with neotectonic movements. It is noteworthy that results obtained using the PSInSAR method have low coherence in the areas of vertical displacements detected by using the DInSAR method. Big and intensive vertical displacements are not detected while using the PSInSAR method. Only using both methods (DInSAR and PSInSAR) allows to generate a complex map of vertical displacements, which is essential for the realization of local zoning plans in the USCB area. 

The analysis of archival SAR data recorded between years 1996 and 1999 by ERS2 satellite is presented. The data was provided by the European Space Agency. The analysis was conducted for radar images of Upper Silesian area located in southern Poland. One of the largest stone coal deposits in Europe are located in this area. The exploitation of coal in this region has continued for over two centuries. In the years of taking the radar images a lot of mines in this area were closed and the others enlarged the coal production. Most of mines are located beneath the highly industrialized and densely inhabited  regions.

An intensive and long lasting underground exploitation has resulted in numerous surface deformations. They were usually of subsidence type and created various size subsiding troughs. Their development, distribution and sizes were determined by the volume of excavated coal and the depth and location of mining works.

In the analysis all the ERS2 radar scenes available for Upper Silesia area in the objective time period were processed. To create the interferograms we have used pairs of images of appropriate coherency, baseline value and time delay. As a result, the twelve interferograms with maximal baseline of 200m and time delay of 140 days were created and used for further analysis. To locate the detected depressions precisely, the phase unwrapping was applied to interferograms. As a result, several fringes of ellipsoidal shape were detected in selected mining areas. They were identified as surface troughs of various size and amplitude. Comparing the successive interferograms we have detected the slight trough migration in the analysed time period. It makes a possible to track the development of subsidence process in time. It also allows us to correlate the seismic emission induced by mining activity with subsidence process.

As an example we discuss in detail the migration of SAR detected subsiding troughs in vicinity of Bytom town. We correlated the process of troughs evolution with mining works in this area especially with the depth and location of excavations and seismic activity.

We present surface displacement results for urban infrastructure monitoring applications derived from  Interferometric Synthetic Aperture Radar (InSAR). Our analysis uses a novel InSAR method, Homogenous Distributed Scatterer (HDS)-InSAR that exploits both persistent point and distributed scatterers by using adaptive multi-looking of statistically homogenous pixel neighbourhoods. The concept is to average a pixel only with neighbours presenting similar scattering properties, thus preserving the boundaries between areas representing different types of backscatter.  The adaptive filtering reduces the noise of smooth, low-backscatter areas such as asphalt and bare ground, while optimally preserving the spatial resolution to increase both the accuracy and geographic precision of the deformation result. The HDS method requires roughly the same number of scenes as Persistent Scatterer Interferometry (PSI), that is typically 15 scenes or more for robust deformation results, while increasing point densities by an order of magnitude over PSI.

Medium-to-high resolution RADARSAT-2 image stacks were acquired over the Canadian cities of Vancouver, Winnipeg, Regina, and Montreal and a TerraSAR-X stripmap stack over Montreal to capture representative examples of urban infrastructure including buildings, bridges, tarmac, and roads. Many cities have a variety of aging infrastructure where due to limited resources, not all infrastructure can be surveyed on a regular basis, and it is envisioned that InSAR would enable a more targeted monitoring process to help identify infrastructure in greatest need of repair. HDS-InSAR results are promising, particularly using high-resolution spotlight data to detect seasonal deformation and potentially hazardous long-term deformation trends of urban infrastructure at high accuracies and dense coverage.

An example of high-resolution linear deformation rate maps over the Vancouver international airport from RADARSAT-2 data generated using Differential InSAR (DInSAR), PSI and HDS-InSAR methods is shown in Figure 1.  Green areas are stable while subsiding tarmac and buildings are yellow and red. While DInSAR smears information over large areas leading to an averaging of deformation information between tarmac, buildings, and surroundings, HDS can clearly differentiate between changes on the tarmac and neighbouring buildings to provide a more accurate assessment for deformation monitoring.  Compared to the PSI result, the HDS-InSAR result has significantly increased spatial coverage, capturing deformation of the airport buildings and subsiding tarmac that are missed by PSI. 

Filtering methods were also developed to identify temporal patterns that can be associated with risk to above- or below-ground infrastructure.  Of particular importance is the separation of temperature-correlated displacement (estimated using meteorological data) that is anticipated in infrastructure design from long-term displacement trends which may pose a hazard to infrastructure.  Examples provided include aging bridge infrastructure in Montreal that exhibits strong thermally-correlated displacement superimposed on a much weaker long-term deformation signal.  Examples are also presented over Regina, where it is found that although long-term linear displacement and seasonal displacement is benign across most of the city, some areas such as the airport tarmac and several major roads experience strong seasonal displacement consistent with frost heave that can result in significant damage to asphalt.

HDS-InSAR combined with temporal matched filtering allow wide areas of infrastructure to be remotely monitored on a regular basis and enables a more targeted monitoring process to help identify infrastructure in greatest need of repair.

    Planning future seismic acquisitions in remote regions requires a non-destructive method to demonstrate the subsurface properties and structural framework. Mapping deformation patterns on the surface which are controlled by the underlying subsurface structure, might also be an indication for subsurface parameters which could determine the next seismic survey location. Taking this idea further, we present the recent outcomes of our research which is conducted over a region with diverse anthropogenic activities but fairly simple geological conditions.

    Over large mining areas, dynamic behavior of a reservoir due to volume or pressure changes may cause significant surface deformation. To estimate the deformation rate including both downward and upward motions, traditional geodetic measurements gradually are swapped with Differential Interferometric SAR (DInSAR) remote sensing techniques to provide very precise displacement maps along the satellite line-of-sight (LOS).

    The current research is undertaken across the Surat Basin, Australia in which for the nature of agricultural and coal mining operations at shallow depth (~300m to 700m), the surface deformation has been observed since 2007 to present day. The advanced multi-image DInSAR processing technique , so called Small Baseline Subset (SBAS) was conducted for both C-band RADARSAT-2 and L-band ALOS PALSAR descending and ascending data with different temporal and spatial baselines. Initially, interferogram stacking results suggested that among four major ALOS PALSAR scenes which cover the proposed area in Surat Basin, two scenes have significant deformation signal patterns after 2007. Based on stacking results, the SBAS processing technique that solves for deformation rate and the residual topographic noise simultaneously was used for coherent (i.e. more than 0.3) interferograms of ALOS scenes which contained the predetermined deformation signals. Time series reconstruction of linear deformation rates in LOS direction, shows four hotspots with significant ground settlement and uplift starting in 2008 which relates to gradual and localized deformation. The localization of these four candidates with surface deformation was also confirmed by implementing SBAS code for descending RADARSAT2 images after 2012 to present day. It is sensible to mention that the rock formation in areas with subsidence is not different from the area with uplift, showing that the sedimentary rocks and young-aged alluvium are not the controlling factors for the deformation signals detected by DInSAR but other man-made interactions can be responsible. 

Based on the resultant deformation footprints after advanced DInSAR processing, the subsurface structural simulation using in-situ geological characteristics, well logging, seismic interpretation, along with interferometric observations is the next objective to achieve to project the properties of shallow depth as well as the strain history in relationship with controlling geological structures. 

A road embankment in Northern Italy, made up of reinforced earth, was affected, in 2013, by severe instability problems. Following huge rainfalls, large cracks appeared on the National and Provincial relevance roads crossing the structure. For public safety, the area was closed causing a serious impact on the local traffic.

In order to assess, in a very short time, the area involved in the phenomenon and its deformation features, a Terrestrial SAR Interferometric (TInSAR) platform was quickly realized in front of the structure thus allowing an overall view on the area of interest. An IBIS-L equipment (by IDS S.p.A.) was installed by a non-invasive quick-installation basement (QUIB^TM) at a distance from the scenario ranging from about 200 m and 400 m. In a few hours, the system started collecting data in a continuous mode with a sampling rate of about 5 minutes and a spatial resolution of 0,75 m (in the range direction) and 4,5 mrad in the cross-range direction. The platform was also equipped with a satellite remote control and data transfer system thus allowing the continuous monitoring for early warning purpose from the data processing centre in Rome.

In less then a couple of days the TInSAR monitoring system allowed to identify the boundary of the unstable area and to quantify the displacements velocity affecting the structure (up to 2 mm/day). The recorded displacement was confined to the embankment area, while no movements were recorded in the surrounding area. Specifically, a fairly homogeneous trend of movement was recorded over the entire embankment, thus suggesting for a rigid movement towards the valley with limited internal deformation.

The integration of TInSAR data with data collected by other monitoring systems (i.e. topographic surveys, inclinometers) and with information inferred by geological and geotechnical site investigations allowed to  achieve a preliminary interpretation of the instability process. Based on these results, the design of stabilization countermeasures was done.

However, in view of the massive earthworks taking place in the instable area lasting some months with the involvement of several workers in an unsafe area, the TInSAR monitoring system was used as an emergency solution to provide real time control and, eventually, for alert purposes. 

During the 17 months of TInSAR survey, lasted from the emergency phase up to the end of remediation works, localized accelerations were detected (mainly related to huge rainfalls) and promptly reported to the authority in charge of the emergency management. However, none of such phenomena was assessed to be directly related to the overall dynamic evolution of the structure.

As a matter of fact, following the initial values (up to 2 mm/day), a gradual decrease of the displacement velocity was detected up to an overall stabilization of the embankment, thus allowing to the safe reopening of the roads to the local traffic.

Furthermore, it is worth noticing that a Terrestrial Laser Scanner survey was also performed by a Riegl VZ1000 sensor in order to get the real colour high resolution point cloud of the structure and for the georeferencing of TInSAR data on a 3D model. This integration allowed for a rigorous localization of the detected local instabilities of the slope and for a suitable data interpretation. 

Greece lies at the convergent boundary of the African and Eurasian plates. On a  local scale the tectonics   become complex with the existence of microplates such as the Aegean plate-extending over the central part of the Aegean sea-interacting with the Anatolian plate to the east. The subduction of the African plate underneath the Eurasian plate is occurring along the island of Crete in Greece, and it drives directly (plate interface) or indirectly (upper plate) the seismicity of the region.

The boundary between the Eurasian plate and the African plate is widely referred to as the Hellenic Arc. It is an arcuate tectonic feature of the Eastern Mediterranean Sea related to the subduction. It consists of an oceanic trench (Hellenic Trench), one volcanic arc (Aegean Volcanic Arc) and an inner non-volcanic arc. Two more ocean floor features define the tectonic processes of the region. The Hellenic Trench is a signature of the so-called forearc basin, the area between the volcanic arc and the surface expression of the subduction (Mediterranean ridge) characterized by lack of sediment deposition. Major characteristics of the subduction include E-W extension on the edges of its 1200km length, variability in seismicity and N-S extension in the bac-arc area.

Crete is part of the non-volcanic arc that includes a number of other islands such as Rhodes and Karpathos to the east and the Ionian islands to the west. This region is characterized by high seismicity (highest in Europe), it is capable of producing M8+ earthquakes, has been the source of great earthquakes in the past (i.e. 365 AD western Crete, 1303AD eastern Crete) and is therefore of paramount importance in terms of seismic hazard. Vertical deformation observed on the island of Crete is primarily due to uplift associated with the plate convergence (Ganas and Parsons, 2009) while its horizontal motion towards south to southwest direction is due to the interaction of the micro plates with the Eurasian and African plates.

In this work we present ground velocities for the entire island of Crete as obtained from the analysis of Synthetic Aperture Radar (SAR) imagery. In particular, deformation has been measured using eight (8) frames providing a complete geographic coverage of the island corresponding to more than 200 of ERS-1, ERS-2 and ASAR/ENVISAT acquisitions from 1992 until 2010. The data have been processed with the Small BAseline Subset technique that estimates velocities associated with continuous deformation, such as those typical of motions near subduction zones. Along with the observed large scale deformation regime, associated with the subduction zone, local displacement patterns are also revealed, attributed mainly to soil erosion phenomena and anthropogenic processes.  

Because of the proximity of the island to the subduction zone, and the complexity of regional tectonics, deformation shows a large variability along the island both in the N-S and E-W directions. We compare our results with available geodetic data from permanent GPS stations of the NOANET network and other relevant studies in addition to simple geodynamic models. The latter consists of a mathematical model/representation of the plates’ interaction, also used for the interpretation of the results.

This work was funded by the European Union 7th Framework Programme (FP7-REGPOT-2012-2013-1), in the framework of the project BEYOND “Building a Center of Excellence for EO-based Monitoring of Natural Disasters”, G. A. 316210.

Volcán de Colima is a steep sloped explosive stratovolcano located in southern Central Mexico, and one of the most active volcanoes in North America. Major recent historical eruptions occured in 1818 (VEI 4) and 1913 (VEI 5), which removed several hundred meters from the summit of the volcanic edifice [1], and point towards the activity being marked by 100-year cycles which terminate in a large Plinian eruption. Also, five large flank collapse events have been identified during the Holocene [2]. Since the beginning of the most recent eruptiove period in 1998, the type of activity has been varying between predominantly explosive, dome building and dome collapse.

Between 2007 and 2011, the activity at Colima was characterized by dome extrusion. The volcano then entered a period of low activity, which lasted until January 2013, when a series of explosions took place which initiated a new, still ongoing period of dome growth. The historical eruption of 1913, as well as the renewal of the activity in 2013, were both preceeded by longer periods of low activity, and only very limited short term precursors. The year 2012 at Volcán de Colima is therefore a good example to study volcanic activity in periods of quiescence, but leading up to an eruption. Furthermore, the possibility of a larger event in the future make it a particularly important volcano to study.

We have acquired TerraSAR-X data in spotlight mode for ascending and descending tracks over  Colima, obtaining a high spatial resolution of up to 2 m, and a temporal resolution of up to 11 days. Here we present the time series of the dome deformation between February and December 2012.

We generated interferograms using an updated version of DORIS, accounting for the doppler variation in along track direction [3] and subsequently analysed the time series of the deformation pattern with the small baseline – persistent scatterer (PS) approach implemented in the StaMPS software. We removed the topographically correlated atmospheric contribution by applying a linear correction to each small baseline interferogram. In combination with a high resolution LIDAR digital elevation model the spotlight data allows the detection and quantification of slow deformation in the region of the dome in an unprecedented spatial and temporal resolution, considerably higher than those achieved by other methods.

The results from the PS approach are compared to those from processing the interferograms using the GAMMA software, and deriving the time series from the filtered, unwrapped and geocoded interferograms using GIAnT. We find that, in this scenario, using GIAnT offers disctinct advanages, in particular the faster processing time and overall spatially smoother results.

The velocities in either look direction reach up to 10cm/year in line of sight, the maximum subsidence being directed towards the centre of the dome, but very low velocities at its borders. The velocities in the center of the dome strongly decrease over the time period analysed, with only a fraction of the total displacements occuring the last months of 2012. We explore whether this pattern may be explained with cooling, the collapse of bubbles in the viscous dome material, or even presents a first sign for the looming renewal of volcanic activity at the beginning of 2013 [4].

 

 

1. González, M.B., J.J. Ramı́rez, and C. Navarro, Summary of the historical eruptive activity of Volcán De Colima, Mexico 1519–2000. Journal of Volcanology and Geothermal Research, 2002. 117(1–2): p. 21-46.

2. Borselli, L., et al., Flank collapse scenarios at Volcán de Colima, Mexico: A relative instability analysis. Journal of Volcanology and Geothermal Research, 2011. 208(1–2): p. 51-65.

3. Jendryke, M., et al., Using Open-Source Components to Process Interferometric TerraSAR-X Spotlight Data. International Journal of Antennas and Propagation, 2013. 2013: p. 13.

4. Salzer, J.T., et al., Satellite radar data reveal short-term pre-explosive displacements and a complex conduit system at Volcán de Colima, Mexico. Frontiers in Earth Science, 2014. 2.

 

The world's 1500 active subaerial volcanoes pose a diverse set of hazards to both people living in their vicinity and the global community. Advanced warnings of the possibility of an eruption can mitigate these risks and are currently provided by several approaches across different disciplines. Due to the cost of expanding many of these methods to cover all 1500 volcanoes, satellite based geodesy provides a unique opportunity for routine global volcanic monitoring.

Currently, technical limitations of operative satellites greatly reduce their ability to perform routine global monitoring. Satellites with revisit times of around a month may be unlikely to image pre-eruptive deformation lasting weeks to months, whist sporadic image acquisition also hinders routine monitoring. However, the Sentinel-1 satellites provide a unique opportunity to overcome these technical difficulties, and consequently expand routine monitoring to most of the world's 1500 active volcanoes. Low revisit times (of around 6 days) allow for multiple interferograms of month long pre-eruptive deformation, whilst use of the European Data Relay System will allow for routine global data acquisition. 

We present here initial progress on an algorithm to identify changes to the background deformation signal extracted from the large number of interferograms generated in the course of routine global volcano monitoring using Sentinel-1.  This will entail separation of new deformation signals from those due to other factors, such as seasonal or steady state deformation. 

The algorithm will be designed to output probabilistic information regarding deformation and any potential magma movement. Tropospheric contributions to the phase delay are likely to be the main source of error, and atmospheric models will be implemented to reduce this error source.  In addition, filtering of repeated acquisitions will also allow reduction in tropospheric contributions producing more tightly constrained statements regarding deformation.  Deformation due to other causes (e.g. seasonal) is also likely to feature in interferograms, and we aim to identify and remove this through the running of the algorithm over the back catalogue of data.

The Mattervalley in the Swiss Alps stretches from the bottom of the Rhonevalley to the famous station of Zermatt. To both sides of this valley there are numerous unstable slopes which can cause damage to inhabited areas, tourist installations, roads and railways. Many landslides, rockslides, and rockfall-sources are active and the annual costs for mitigation and countermeasures are high. Due to the difficulties in the detection of the active, slow or dormant portions of the landslides, the use of satellite SAR interferometry can positively impact the current practices used by the local authorities to plan and implement hazard mitigation activities. In order to estimate the progressive deformation of the terrain at millimetre accuracy, multi-temporal InSAR methods (Persistent Scatterer Interferometry (PSI) and small-baseline approaches) were performed using stacks of ERS-1/2, ENVISAT, TerraSAR-X, Cosmo-SkyMed and Radarsat-2 SAR data acquired between 1995 and 2013. In mountainous regions the number of coherent scatterers is generally limited by the sparse urbanization, the large forest cover, and areas of shadow and layover, but a significant improvement was possible in recent years with the use of SAR sensors characterized by a higher spatial resolution and a shorter revisiting time. Investigations performed at altitudes ranging from about 1'200 m to more than 4'000 m a.s.l., where considerable parts of the slopes are without any vegetation or dominated by scarce patches of alpine grass, well indicate how landslides with movement rates up to 30-40 cm/year can be monitored. But the investigated area also encompasses numerous active rockglaciers with deformation rates up to 10 m/year and more. Indeed, the velocity of rock glaciers has risen significantly since the 1980s over the whole alpine range in response to a significant permafrost warming and recently destabilized (or “surging”) rockglaciers showing morphological indices of landslide-like mass wasting have been reported from several regions, including Grosse Grabe, Breithorn, Dirru and Grabengufer in the Mattervalley. The activity of these very rapidly moving rockglaciers cannot be thoroughly quantified with multi-temporal InSAR methods via phase unwrapping, because interferograms with short perpendicular baselines from Cosmo-SkyMed, TerraSAR-X, Radarsat-2, ENVISAT ASAR and JERS-1 show all significant signal decorrelation after the 4, 11, 24, 35 and 44 days repeat intervals of these SAR sensors operating at X-, C- and L-band, respectively.

A terrestrial radar interferometer compliments satellite SAR data in time and space, allowing the measurement of additional displacement vectors and velocity classes. The GAMMA Portable Radar Interferometer (GPRI) can operate over distances up to about 10 km with resolutions of 0.75 m in range and of 6.8 m in azimuth at 1 km range. The measurement accuracy is less better 1mm at 1 km and depends on variability of the atmospheric water vapour. Scans of even more than 180º size can be taken in less than one minute to allow for near real-time monitoring. GPRI measurements of the Grosse Grabe, Breithorn, Dirru and Grabengufer rockglaciers were performed from a position on the opposite side of the valley during campaigns lasting a few days every summer since 2010. The distance to the target was considerably large, from about 3 to more than 6 km, with significant atmospheric artefacts. Scans performed with a 10 minutes interval were interfere in series to compute cumulative displacement maps after compensation of the atmospheric artefacts. The highest movement rates were measured for the Grabengufer rockglacier, with values on the order of 10 cm/day, which is in agreement with measurements performed with other surveying techniques like GPS and total-station. The three other active rockglaciers of our test area, i.e. Grosse Grabe, Breithorn and Dirru, were characterized by movement rates at the front of more than 1 cm/day. In our contribution we present the processing strategy, including in particular the characterization and mitigation of atmospheric artefacts, and discuss the results.

Introduction

Persistent Scatterer Interferometry (PSI) is an advanced technique to extract accurate and dense data on ground movements from time series of Synthetic Aperture Radar (SAR) images. In this study PSI is utilised to measure the spatial extent of subsidence in the upper Scheldt estuary. Particular attention is paid to the identification of built heritage exposed to these ground movements.

Study area

The port of Antwerp is situated along the upper Scheldt estuary in the north of Belgium (Figure 1a). Owing to its geological setting this region is prone to ground subsidence and flooding, while considerable economic and cultural assets are potentially at risk. In general, the main causes of ground subsidence in alluvial plains and deltas include: compaction of young sediments, oxidation of organic layers related to drainage, exploitation of (deep) natural resources and tectonic deformation (Panel on Land Subsidence 1991; De Lange & Gunnik 2011). In addition to direct damages to buildings and infrastructure, subsidence may cause long-term impacts on the operation of drainage systems and flood frequency.

The Scheldt estuary NW of Antwerp consists of a typical polder landscape, largely lying below mean high tide level (between +0.8 and +4.5 m TAW). After the last glacial, due to the rising sea-level, the Scheldt valley was gradually filled up with fluvial and tidal sediments (Figure 1b). After c. 5700 BP significant peat accumulation occurred in the entire area (Kiden 2006). Around 0-600 AD the peat growth stopped as the area became subject to tidal influence. The tidal amplitude increased considerably after c. 1100 AD, when a new connection to the sea through the Western Scheldt was established and thick layers of estuarine sands and clays were deposited on top of the peat. From the 13^th century onwards dikes were constructed (Mijs 1973), preventing further sedimentation in the newly reclaimed lands. Still, on several occasions during the following centuries, dikes were breached and large tracts of land inundated during storm surges and military conflicts. Since the early 19^th century many of the polders were claimed by the expanding Port of Antwerp and the ground surface artificially raised by thick layers of land fill material. The areas bordering the Scheldt valley, including the historic city centre of Antwerp, are covered by Pleistocene sands overlying Tertiary units.

Data sets

In this case study all available SAR data from the ERS1/2 and ENVISAT missions covering the area of interest were used, providing two data series of 67 (ERS track 423, dated between 1992-2001) and 74 (ENVISAT track 423, dated between 2003-2010) images, respectively. Both data series were processed using the PSI technique: multiple differential interferograms were generated, coherent pixels identified based on numerical and statistical analyses, and high-precision displacements estimated (Ferretti et al. 2001). For further analysis the datasets were reduced to an area of 50 x 32 km around the city and port of Antwerp and only including points with coherence values greater than 0.70. The final subsets include annual average velocities (in mm/year) of 35,605 (ERS) and 4,947 (ENVISAT) permanent scatterers, corresponding to average point densities of 22.3 and 3.1 points/km² (Figures 1c-d)

Results

Preliminary interpretation of the PSI results, combined with geological and geotechnical data, shows a clear pattern of subsidence in the Scheldt polders, whereas in adjacent zones the ground is relatively stable or even slightly uplifting. Cross-tabulation of the PSI data with the Quaternary geological map indicates continuous subsidence in the polder area (map unit “Alluvial + tidal” in Figure 1b), with maximum values of -7.34 (ERS) and -7.40 (ENVISAT) mm/yr (Figures 2a-b). Average observed displacements in the Pleistocene sand belt are +0.23 (ERS) and +0.78 (ENVISAT) mm/yr (Figures 2c-d). Most likely the observed subsidence can be attributed to the compaction of alluvial clay and peat layers and the presence of a considerable land fill cover in urban areas. Further research will focus on establishing a quantitative relation between these parameters.

Concerning the built heritage of Antwerp, the highest risk for instabilities due to ground movements can be expected in the dock areas and in the urban quarters on the left bank of the Scheldt, both situated in zones of subsidence (Figure 3). As the historic city centre is built on relatively stable ground, the risk for its many monuments here seems limited.

Acknowledgements

This work is carried out within the framework of the GEPATAR (GEotechnical and Patrimonial Archives Toolbox for ARchitectural conservation in Belgium) project financed by BRAIN (Belgian Research Action through Interdisciplinary Networks). GEPATAR aims to improve the risk management of the federal patrimony

References

De Lange, G. & Gunnik, J. L. (2011). Bodemdalingskaarten: Concept-rapport. Deltares, Utrecht, The Netherlands, 16 pp.

Ferretti, A., Prati, C. & Rocca, F. (2001). Permanent scatterers in SAR interferometry. IEEE Transactions on Geoscience and Remote Sensing, 39 (1), pp. 8-20.

Kiden, P. (2006). De evolutie van de Beneden-Schelde in België en Zuidwest-Nederland na de laatste ijstijd. Belgisch Tijdschrift voor Geografie, 3, pp. 279-294.

Mijs, M. (1973). De landschapsgeschiedenis van de Scheldepolders ten noorden van Antwerpen. Bijdrage tot de historische geografie van de Scheldepolders. Tijdschrift van de Belgische Vereniging voor Aardrijkskundige Studies, 42, pp. 39-124.

Panel on Land Subsidence (1991). Mitigating Losses from Land Subsidence in the United States. The National Academies Press, Washington D.C., 68 pp.

Relationships between tectonic and magmatic activity in the Central American Volcanic Arc (CAVA) are derived from temporal and spatial patterns of seismicity, eruptive activity, and surface deformation.  Interferometric Synthetic Aperture Radar (InSAR) surveys spanning from 2006 to 2010 detected little evidence of surface deformation at volcanoes along the CAVA. This may be attributable to the inherent difficulty of applying InSAR analysis in this region due to dense tropical vegetation, steep topographic slopes, and significant atmospheric effects. Quantified surface deformation can be used in models to better constrain timing and facilitation of volcano tectonic interactions. To better identify volcanic surface deformation, a multidisciplinary approach is applied to overcome issues associated with the application of InSAR in Central America. More focused remote sensing studies using synergistically sophisticated InSAR-derived methods (e.g., time-series approaches), new SAR sensors (e.g., RADARSAT-2, UAVSAR, COSMO-Skymed, Sentinel-1) and optical imagery (e.g., Landsat, ASTER) are applied to recent seismic and volcanic events.  Remote sensing analysis combined with stress transfer modeling help to further understand possible feedbacks between tectonic and volcanic activity.

Earthquake swarms at Momotombo volcano in June 2000 and Cosigüina volcano in September 2002 exhibit timing and magnitude indicative of magmatically induced seismicity.  Magmatic events strong enough to incur intermediate magnitude (~Mw 5) shallow earthquakes would be expected to generate surface deformation signals. Available ERS-2 data covering these earthquake swarms will be processed, and resulting surface deformation will be used to model possible magmatic activity. At the northern end of the CAVA, Pacaya Volcano in Guatemala has been erupting continuously since March 2013.  Flank instability is prevalent at Pacaya, including landslides from the Pleistocene that reach 25-kilometers in length.  Other smaller mass wasting events have occurred as recently as 2010.  Aided by the removal of vegetation from recent activity, new RADARSAT-2 data acquired from June 2014 can produce improved measurements of surface deformation and provide the ability to model ongoing magmatic and flank instability processes.

Seismogenic fault distribution of low magnitude (

Multidisciplinary investigations using SAR, optical imagery, GPS, and stress analysis models are capable of characterizing surface deformation of volcanoes along the CAVA.  Case studies of earthquake swarms at Momotombo and Cosigüina, ongoing deformation at Pacaya, and inter-volcanic seismic activity near Momotombo can result in well-defined local surface deformation. Surface deformation measurements are the primary data source for modeling subsurface magmatic activity.  Magmatic models are subsequently used to model Coulomb failure stress changes in the upper crust.  Correlation of modeled stress changes with observed seismic activity identify likely sources of stress transfer causing destructive upper crustal earthquakes in Central America. A better understanding of how volcanic eruptions are triggered via earthquakes, or vice versa, significantly increases the potential for hazard assessment in Central America.

The application of SAR interferometry (InSAR) has been widely used to monitor volcanic-induced deformation in active areas. Recently, the Japan Aerospace Exploration Agency (JAXA) successfully acquired data from the Phased Array Type L-band Aperture Rader-2 (PALSAR-2) aboard the Advanced Land Observing Satellite-2 “DAICHI-2” (ALOS-2). ALOS-2 is the follow-on JAXA L-bad SAR satellite mission of ALOS and provides L-band SAR data. Using the PALSAR and PALSAR-2 data, we will present the application of InSAR to several types of volcano deformation. We will show preliminary analysis and modeling for Kuchinoerabujima, Japan.

 

Kuchinoerabujima is an active volcanic island located on the volcanic front of the Ryukyu island arc, off Japan’s SW coast. On August 3, 2014, the Japan Meteorological Agency (JMA) reported that a new eruption occurred at Kuchinoerabujima in the vicinity of Shin-dake crater. Until this event, eruption was not observed for more than 30 years; however, seismic swarms were accompanied by a radial outward pattern from the summit crater repeated almost every two years since 1999. Recent eruptive activities of Kuchinoerabujima volcano occurred at two active craters of Shin-dake and Furu-dake. Ground displacements near the summit area of Shin-dake were also detected by conventional InSAR analysis using PALSAR data. We applied the InSAR time-series analysis to PALSAR data acquired on both ascending and descending orbits from May 2006 to March 2011. This analysis identified enough coherent pixels to successfully resolve the spatial and temporal deformation. The line-of-sight (LOS) displacements show a rather complicated pattern compared with previous results obtained using GPS measurements and InSAR analysis. The mean velocity maps show two focused areas of LOS shortening located beneath Shin-dake and Furu-dake at a rate of 20 mm/year, confirming the inflation trend. The observed deformation near the summit area of Shin-dake was consistent with previous results. Also, it suggests another deformation source beneath Furu-dake, which was not clearly accounted for previously. We model the Kuchinoerabujima volcano sources that produced clear and distinct fringe patterns using a Markov Chain Monte Carlo optimization. Two regions of inflation were modeled by two point sources located at depths of 0.5 and 0.3 km and with volume changes of 1.7 x 10⁴ and 6.3 x 10³ m³ respectively. The location, depth, and volume change of a point source beneath Shin-dake is consistent with that of inflation and demagnetization pat the summit area of Shin-dake, indicating expansion and heating of shallow region at about a few hundred meters beneath the Shin-dake summit. We will present conventional InSAR results using PALSAR-2.

 

The viability of using L-band observations as a monitoring tool was established PALSAR data. Longer-term continuous monitoring with new SAR missions, such as SENTINEL-1 and ALOS-2, will provide us with more opportunities for new insights into volcanic processes.

On 12^th May 2008, a magnitude 8.0 earthquake happened in Sichuan province, China. It was reported that 60000 landslides and rock collapsed were trigged by the great earthquake over mountain area. Wenjiagou landslide is one of the biggest landslides triggered by the Wenchuan great earthquake. After the landslide event, mass debris are deposited in the gully of Wenjiagou, which have become potential threaten to the safety of local people. Therefore it is urgently needed to monitor the deformation information of the landslide.

Synthetic aperture radar interferometry (InSAR) is a powerful technology for measuring ground deformation by analysing the phase difference at different times. However, applications of DInSAR are limited by some drawbacks such as spatial and temporal decorrelation, atmosphere disturbance delay, and DEM error, especially at mountain areas  In order to overcome these limitations, several advanced techniques have been proposed, such as Persistent Scatterer Interferometry (PSI) and Small baseline Subset Algorithm (SBAS). However, those above methods have some limitations in landslide monitoring due to low density of PSs or coherent points and their inhomogeneous distribution. To improve the spatial density of the measurement points, Distributed Scatterers Interferometry (DSI) has been proposed to monitor the surface deformation of large scale suburb areas

In this study, we try to monitor the deformation of the giant landslide at Wenjiagou after Wenchuan earthquake using time-series InSAR technique on DSs. TDX and TSX bistatic SAR data acquired on Feb. 1, 2013 are processed to generate high precision DEM of the post-landslide in this work (no external DEM is used). DSs can be identified using both classified information and statistical characteristics. A series of Radarsat-2 HH polarization images collected in Wenjiagou from 2014.6 to 2014.9 are used to get the deformation of the giant landslide. Experimental result shows that the landslide moves during the observation.

   Underground coal mining often causes persistent land subsidence, which would lead to huge threats to infrastructure and safety in the coal area. Traditional levelling and GPS methods are able to produce reliable measurement of ground subsidence. However, these field surveys are time-consuming and cannot provide deformation map with high spatial sampling density.

         Synthetic aperture radar (SAR) interferometry (InSAR) is a powerful technology, which can obtain high precise elevation and surface deformation information along line of sight (LOS) using phase information from different SAR images. However, conventional InSAR usually suffers from temporal, geometrical and atmospheric decorrelation. In order to overcome these limitations, several advanced techniques have been proposed, such as Persistent Scatterer Interferometry (PSI) and Small baseline Subset Algorithm (SBAS). In recent years, Distributed Scatterers Interferometry (DSI) has been proposed to monitor the surface deformation with high spatial density of measurement points. The main drawback of DSI is the computational efficiency in large scale suburb areas.

         This paper presents an approach to obtain ground deformation using time-series InSAR technique combining PSs and DSs. PSs are selected based on their coherence stability in the stack of interferograms. In order to identify DSs, a selecting strategy combining both classified information and statistical characteristics is used. To control the error propagation and improve the computational efficiency, deformation rate and DEM error on PSs are firstly retrieved using conventional PSI. Then a region grown-based strategy is applied to extract deformation rate on the useful DSs. A series of Radarsat-2 HH polarization images collected in Huainan are processed to verify the effectiveness of the proposed method. A dense ground surface deformation map is obtained. The experimental results show that the potential application of the proposed method.

San Miguel volcano, also known as Volcan Chaparrastique, is a stratovolcano located in the eastern El Salvador. This volcano is one of the most active volcanoes of the country and its summit reaches 2130 m elevation from near sea level and forms a prominent land mark.

On 29^th of December 2013, at 10:30 local time, the Volcan Chaparrastique began its eruptive activity. The eruption was mainly characterised by dense ash clouds and strong gas emission, and the initial explosion caused a plume that rose to an altitude between 5 and10 km. The surrounding areas were covered by the ash fallout, affecting in particular the Chinameca e San Jorge villages, on west of Volcan Chaparrastique.

During this major erupting period, a task force from Istituto Nazionale di Geofisica e Vulcanologia (INGV), were triggered to support on ground monitoring activities, through the installation of additional gas and geodetic instruments. In such a framework, the SAR satellite monitoring was initiated.

In this work we have performed both standard differential SAR interferometry (DinSAR) and multi-temporal techniques analysis. In particular, the Small BAselines Subset (SBAS) method is adopted. The aims is to detect and measure the occurred deformation of the volcano edifice. The inferred displacement is also analysed considering the ongoing seismic activity.

For the purpose of the work, we take the advantage of a series of 35 SAR images acquired by the satellite constellation of the Italian COSMO-SkyMed mission. The exploited high resolution data are all in Stripmap mode, on right ascending viewing angle. The dataset covers a time span of few months in the period from January 2014 to August 2014. Unfortunately, no Stripmap mode image  is available before the beginning of the eruption, therefore, only post-eruptive displacement can be detected. In addition, GPS measurements carried out on a network installed to monitoring the current activity are also used and compared with interferometric results.

Processing satellite C-band SAR imagery with Persistent Scatterers (PS) and Small Baseline Subset (SBAS) techniques over non-urban regions can be challenging due to temporal decorrelation and loss of phase coherence. These techniques can often result in incomplete views of the spatial extent of ground motion occurring in rural regions of interest and, in some cases, may not detect motion at all if the latter occurs outside urbanised, persistently reflective or coherent areas. Such limitations have been proved by the feasibility study recently carried out for Great Britain at the British Geological Survey (BGS), in preparation for the creation of a baseline, nationwide ground deformation product using archive ERS-1/2 and ENVISAT SAR imagery, and towards sustainable exploitation of data from new explorers [Cigna et al. 2014].

We take on the challenge of processing rural regions with C-band InSAR for the South Wales Coalfield, and carry out a proof-of-concept for the exploitation of forthcoming long stacks of C-band SAR imagery, currently being built up by Sentinel-1A since October 2014. This test region is typically covered by rural land covers and only a few major urban areas are present, such as Cardiff and Swansea, and the mining towns found in the bottom of the deeply incised valleys to the east of the coalfield [Bateson et al. 2015]. Ground motion is known to be occurring in this area, relating to mass movements on the steep sided valleys, compression and consolidation of normally-consolidated Quaternary deposits and more recent made ground, as well as surface motions relating to the extraction of coal and abstraction and rebound of groundwater.

We employ both the conventional low-pass SBAS technique and the novel Intermittent SBAS (ISBAS) approach [Sowter et al. 2013], the latter applied to analyse intermittently coherent pixels and extend ground motion results to non-urban land covers. We process 55 ERS-1/2 SAR scenes for the period 1992-1999, made available to BGS by the ESA through the Category-1 project id.13543: ‘Enhancing landslide research and monitoring capability in Great Britain using C-band satellite SAR imagery and change detection, InSAR and Persistent Scatterers techniques’.

ISBAS provides greater spatial coverage and a more complete picture of deformation patterns across the South Wales Coalfield, with over 375,000 natural targets identified, and coverage of ground motion results around seven times higher than conventional SBAS. The improved density allows not only the identification of small areas of motion which fall in the gaps of the SBAS result, but also the edges of displacement areas to be more accurately defined and interpreted. ISBAS depicts many of the suspected ground motions, but also reveals large areas of surface uplift at up to 1 cm/year in the east of the coalfield where there is a history of deep and shallow coal extraction. This uplift relates to the recovery of groundwater levels following the cessation of minewater pumping. We examine the relationships between observed ground motions and their causative factors, in particular the timing of coal extraction and minewater pumping histories of these mining sites. This proof-of-concept paves the way for the exploitation of this advanced processing approach using the forthcoming resource of the Sentinel-1A archive.

REFERENCES

Bateson, L, Cigna, F, Boon, D, Sowter, A. (2015). The application of the Intermittent SBAS (ISBAS) InSAR method to the South Wales Coalfield, UK. International Journal of Applied Earth Observation and Geoinformation, 34, 249–257.

Cigna, F, Bateson, L, Jordan, C, Dashwood, C. (2014). Simulating SAR geometric distortions and predicting Persistent Scatterer densities for ERS-1/2 and ENVISAT C-band SAR and InSAR applications: Nationwide feasibility assessment to monitor the landmass of Great Britain with SAR imagery. Remote Sensing of Environment, 152, 441-466.

Sowter, A, Bateson, L, Strange, P, Ambrose, K, Syafiudin, M. (2013). DInSAR estimation of land motion using intermittent coherence with application to the South Derbyshire and Leicestershire coalfield. Remote Sensing Letters, 4, 979-987.

Albanian Preadriatic Depression zone is characterised by a combination of sandy beaches and hilly formations of friable sandstone during the Adriatic coast. This area has experienced significant coastline changes, sea transgression and landslides that have impacted the human activities. We have used the satellite imagery to analyse variations of the Preadriatic Depression and in this paper results of SAR interferometry are presented. The use of the interferometry was done for two purposes - to evaluate vertical movements of the shoreline ground and identify potential landslide areas. We used Envisat and ERS2 images retrieved from ESA using EOLI software to pick up interferometric pairs. The processing of images was done with RAT and NEST software packages. Interferometry results presented in the paper, combined with other remote sensing data and field observations, clearly show the geomorphological and environmental complexity of this tectonic area, as well as the necessity for further multidisciplinary studies and continuous monitoring. During the whole Adriatic shore there are identified shifts of rivers deltas combined with sea transgresion due to beach erosion and subsidence, with a number of buildings now destructed by the sea. Intense vegetation variations and human activities in lowlands impact the quality of interferograms especially in the southern half of Preadriatic Depression. Interferometric fringes correlated with good coherence are more visible in the northern half, strictly correlated with the highland ranges. Use of interferograms for prediction of landslides was undertaken only for a section of inhabitated hilly range characterised by frequent slides, with the goal to pave the way for a full coverage of the whole territory. Fringes were identified in known massive landslide areas, but the noisy background and the relative small size of slides compared with the resolution of SAR images makes difficult the prediction of new landslides; specific analysis based on predefined points of reference were used for further study. This project was launched in framework of National Programme for R&D 2010-2013. 

The research aims to combine deformation information acquired from satellite InSAR and ground-based radar methods in order to improve the understanding of the mechanisms and kinematics of slope processes on the southwest-facing rock slope of Ádjet mountain (Troms county, North Norway).

Ádjet is a 8-km long mountain ridge situated in the northeast side of the glacial formed Skibotn valley, northern Norway. The area is part of the Reisa nappe complex, originally sedimentary rocks trusted and metamorphosed during the Caledonian orogeny. It consists in mica schist and meta-arkose (sandstone) and is affected by several faults and fractures (Boyd et al. 1985; Quenardel, et al. 2008). The regional permafrost limit has been documented to descend from about 990 m.a.s.l. in the western part of northern Norway to about 550 m.a.s.l. in the eastern part (Christiansen et al., 2010). With a ridge up to 1360 m.a.s.l., the upper part of Ádjet is expected to be located in the regional permafrost zone.

The highly fractured material in the source area is transported downslope by various rock slope processes (rockslide, rockfall, rock glacier, debris flow, solifluction), which were not intensively studied until now. Preliminary fieldwork highlights the presence of several lobe-shaped creeping landforms, believed to be rock glaciers, i.e. creeping ice-rich frozen material in cold-climate high-relief regions (Humlum, 2000; Haeberli et al., 2006).

Snow-free TerraSAR-X/TanDEM-X scenes between 2009 and 2014 were processed and highlight rate of displacement of several centimeters/year at various locations along the Ádjet ridge. However, high displacement rates in some areas lead to signal decorrelation. These areas are interpreted as rock glaciers.

A Gamma Portable Radar Interferometer (GPRI) was set up in summer 2014 in the valley, more than 3 km away to the upper part of the slope. The GPRI is a ground-based Ku-band (17.2 GHz) radar system used for deformation mapping and monitoring of unstable slopes (Strozzi et al., 2009; Lowry et al., 2013; Caduff et al., 2014), glaciers (Riesen et al., 2011; Wiesmann et al., 2008) and infrastructures (Werner et al., 2012) up to distance on the order of 10 kilometers. The radar is a real aperture with 2 meters-long antennas, allowing for a ground azimuth resolution of about 7 m at 1 km distance and a range resolution of 0.9 m (3 dB). The radar images are acquired by azimuthally rotating the antennas about a vertical axis. Two receiving antennas have a spatial baseline allowing creation of digital elevation model of the scanned area (Werner et al., 2008). GPRI images were acquired every 5 minutes during 24 days between the 9^th of August and the 1^st of September 2014. These data highlight extremely high velocities up to 11 cm/day on two lobe-shaped landforms, which suggest destabilization processes of rock glaciers in the area.

The use of both satellite and ground-based interferometric observations contributes to a better understanding of kinematics, evolution and relations with external meteorological factors of a fast moving complex in a potential perennially frozen area. It is one of the first research contributions investigating high velocity rock glaciers in high latitude periglacial environment including ground-based radar measurements. This could be especially important in a context of climate changes.

Boyd, R., Mikalsen, Boyd, R., Mikalsen, T., Minsaas, O., Zwaan, K. B. 1985. Geological map Storfjord 1633.4 M 1:50.000, Preliminary version. Norwegian Geological survey.

Caduff, R., Kos, A., Schlunegger, F., McArdell, B. W., Wiesmann, A. 2014. Terrestrial Radar Interferometric Measurement of Hillslope Deformation and Atmospheric Disturbances in the Illgraben Debris-Flow Catchment, Switzerland. IEEE Geoscience and Remote Sensing Letters, 11(2), 434-438.

Christiansen, H. H., Etzelmüller, B., Isaksen, K., Juliussen, H., Farbrot, H., Humlum, O., Johansson, M., Ingeman-Nielsen, T., Kristensen, L., Hjort, J., Holmlund, P., Sannel, A. B. K., Sigsgaard, C., Åkerman, H. J., Foged, N., Blikra, L. H., Pernosky, M. A., Ødegård, R. S. 2010. The thermal state of permafrost in the nordic area during the international polar year 2007–2009, Permafrost and Periglacial Processes, 21(2), 156–181. DOI:10.1002/ppp.687.

Haeberli, W., Hallet, B., Arenson, L., Elconin, R., Humlum, O., Kääb, A., Kaufmann, V., Ladanyi, B., Matsuoka, N., Springman, S., Vonder Mühll, D. 2006. Permafrost Creep and Rock Glacier Dynamics. Permafrost and Periglacial Processes, 17, 189-214, DOI: 10.1002/ppp.561.

Humlum, O. 2000. The geomorphic significance of rock glaciers: estimates of rock glacier debris volumes and headwall recession rates in West Greenland. Geomorphology, 35(1-2), 41–67. DOI:10.1016/S0169-555X(00)00022-2.

Lowry, B., Gomez, F., Zhou, W., Mooney, M.A., Held, B., Grasmick, J. 2013. High resolution displacement monitoring of a slow velocity landslide using ground based radar interferometry. Engineering Geology, 166, 160-169.

Quenardel, J. M., Zwaan, K. B. 2008. Geological map Manndalen 16331 1:50 000, Preliminary version, Norwegian Geological survey.

Riesen, P., Strozzi, T., Bauder, A., Wiesmann, A., Funk, M. 2011. Short-term surface ice motion variations measured with a ground-based portable real aperture radar interferometer. Journal of Glaciology, 57(201), 53-60.

Strozzi, T., Delaloye, R., Raetzo, H., Wegmüller, U. 2009. Radar interferometric observations of destabilized rockglaciers. Proc. ‘Fringe 2009 Workshope’, Nov. 30 - Dec. 4, 2009, Frascari, Italy. ESA SP-667, March 2010.

Werner, C., Strozzi, T., Wiesmann, Wegmüller, U. 2008. Gamma’s portable radar interferometer. Proc. ‘13th FIG Symposium on Deformation Measurement and Analysis’, May 12-15, 2008, Lisbon, Portugal.

Werner, C. Wiesmann, A., Strozzi, A., Kos, A., Caduff, R., Wegmüller, U. 2012. The GPRI Multi-mode Differential Interferometric Radar for Ground-based Observations. Proc. ‘9^th EUSAR European Conference on Synthetic Aperture Radar’, April 23-26, 2012, Nuremberg, Germany.

Wiesmann, A., Werner, C., Strozzi, T., Wegmüller, U. 2008. Measuring deformation and topography with a portable radar interferometer. Proc. '13th FIG Symposium on Deformation Measurement and Analysis’, May 12-15, 2008, Lisbon, Portugal.

The Big Island of Hawai`i consists five coalesced volcanoes: Hualālai, Mauna Loa, Kilauea, Mauna Kea, Kohala. Mauna Loa, the largest, has erupted 39 times since 1832, with the last eruption in 1984 [1]. The volcano summit hosts the Moku’aweoweo caldera from which two volcanic rift zones radiate: the Northeast Rift Zone (NERZ) and Southwest Rift Zone (SWRZ) [2]. These rifts are the reflection of past dike intrusions combined with instability of the SE flank of the volcano, possibly related to slip along a low angle decollement thrust fault 12-14 km beneath the volcano [3].

Geodetic (InSAR, GPS) [1-3-4] and seismic data [5] have been used to characterize recent periods of unrest at Mauna Loa. InSAR studies spanning the period between 2002 and 2005 suggest a magma chamber 4.7-km depth below the summit, with a radius of 1.1 km, and a dike intrusion at 4 to 8-km depth and 8 km-long [3]. These studies, however, are focused on relatively short-term processes (up to a few years), and a longer-term reconstruction of the volcano’s evolution is lacking.

In this work, we use SAR data, exploiting the SBAS technique [6], to study deformation of Mauna Loa from 2003 to 2014, and we try to relate this overall evolution to that the neighboring Kilauea. We use acquisitions from two satellites: ENVISAT (descending track 200 and ascending track 365), spanning from 2003 to the end of 2010, and COSMO-SkyMed (descending and ascending orbits), spanning from 2012 to the middle of 2014. These data are merged time series data from 24 continuously operating GPS stations, which allows us to calibrate the InSAR SBAS time series.  Results show an overall good agreement between the InSAR and GPS time series. The displacement of each portion of the volcano between 2003-2014 has been thus constrained in detail.

The summit area of Mauna Loa has under gone a long-term inflation from 2003 to 2014, with a peak of about 8 cm of vertical deformation between mid-2004 to mid-2005, especially on the summit. Part of this deformation may be related to the instability of the flank.

We identified several distinct periods with linear deformation behavior. This deformation is now being used to obtain the best-fit parameters for the sources in each of these periods through analytical modeling. Subsequently, we will incorporate these model parameters in a 3D model that employs the finite element method to include complexities like topography and vertical/lateral heterogeneities.

REFERENCES

[1] Casu, F., Lanari, R., Sansosti, E., Poland, M., Miklius, A., Solaro, G., and Tizzani, P., "SBAS-InSAR analysis of surface deformation at Mauna Loa and Kilauea volcanoes in Hawaii," Geoscience and Remote Sensing Symposium, 2009 IEEE International, IGARSS 2009, vol. 4, pp. IV-41 - IV-44, doi: 10.1109/IGARSS.2009.5417600, Jul. 2009.

[2] Walter, T. R., and Amelung, F., "Volcano‐earthquake interaction at Mauna Loa volcano, Hawaii," Journal of Geophysical Research: Solid Earth (1978–2012), vol. 111, no. B5, doi: 10.1029/2005JB003861, May 2006.

[3] Amelung, F., Yun, S. H., Walter, T. R., Segall, P., and Kim, S. W., "Stress control of deep rift intrusion at Mauna Loa volcano, Hawaii," Science, vol. 316, no. 5827, pp. 1026-1030, doi: 10.1126/science.1140035, May 2007.

[4] Pepe, A., Ortiz, A. B., Bonano, M., Lanari, R., Lundgren, P., and Rosen, P. A., "Deformation in Hawaii's volcanoes obtained from a ScanSAR-to-stripmap Small BAseline Subset technique," Geoscience and Remote Sensing Symposium (IGARSS), 2010 IEEE International, pp. 769-772, doi: 10.1109/IGARSS.2010.5651771, Jul. 2010.

[5] Okubo, P. G., and Wolfe, C. J., "Swarms of similar long-period earthquakes in the mantle beneath Mauna Loa Volcano," Journal of Volcanology and Geothermal Research, vol. 178, no. 4, pp. 787-794, doi: 10.1016/j.jvolgeores.2008.09.007, Dec. 2008.

[6] Berardino, P., Fornaro, G., Lanari, R., and Sansosti, E., "A new algorithm for surface deformation monitoring based on small baseline differential SAR interferograms," IEEE Trans. Geoscience and Remote Sensing, vol. 40, no. 11, pp. 2375-2383, doi: 10.1109/TGRS.2002.803792, Nov. 2002.

The Xynthia storm was of of the most destructive (>billions of euros damage) and deadliest (> 25 casualties) in France since the 1999 december storm. We show the state of deformation of the coastline in the area of La Faute s/ mer using ENVISAT images. We show that at the very local scale the subsidence can reach 1cm/yr. 

Located at the northern end of the North China Plain, Beijing is ranked as the 5th most water-stressed city in the world [McDonald et al., 2014], and groundwater is the main water source for industrial, agricultural and household activities. With the rapid urban growth, there have been increasing water demands inBeijing. Recent researches [Jia et al., 2007] revealed that theBeijingregion had been suffering from land subsidence due to over-exploitation of groundwater, and more seriously, the rate and extent of land subsidence show an increasing trend. Land subsidence is a severe geohazard threating the safety of public lives and urban infrastructure. Hence continuous monitoring of land subsidence is critical for predicting potential hazards and designing compensation strategies.

 

Interferometric Synthetic Aperture Radar (InSAR) is known as a powerful tool to map large-scale land subsidence with high spatial-temporal resolution. InSAR time series techniques such as Persistent Scatterer Interferometry (PSI) [Ferretti et al.2000,2001] and Small Baseline Subset (SBAS) [Berardino et al. 2002, Mora et al.2003]can minimise the limitations of traditional InSAR (e.g. spatial and temporal decorrelation and atmospheric effects), and have been demonstrated to be able to map ground displacements with precision comparable to traditional geodetic technique such as leveling [Prati et al.2010, Li et al.2012] .

 

InSAR time series has been employed to investigate land subsidence in Beijing. Gong et al. [2009] explored the application of PSI method to monitor land subsidence inBeijingwith C-band Envisat ASAR data acquired from 2003 to 2006. The results show that the spatial pattern of the subsidence is influenced by Quaternary faults, and the subsidence areas often occur in the clay layer thicker than50m. Li et al. [2013] adopted SBAS method to investigate Beijng's land deformation with two adjacent tracks of C-band Envisat ASAR images, the results show that the maximum subsidence rate arrives at100 mm/yr in several key subsidence areas. Liang et al. [2013] acquired the accumulated crustal deformation and its characteristics in Beijing and surrounding areas by traditional D-InSAR method with L-band SAR data, the results show that the accumulated maximum LOS deformation is about37.6 cmduring the time .Compare with the lower resolutions SAR imagery such as C-band ENVISAT ASAR and L-band ALOS PALSAR data, the recent advanced imaging radar sensor such as the TerraSAR-X SAR system with X-band radar wavelength and shorter repeat interval can provide higher spatial resolution and geometric accuracy imagery, which makes it possible to monitor the land subsidence with finer scale [Wegmuller et al.2010, Liu et al.2010]. Chen et al.[2013] utilized C-band Envisat ASAR data and X-band COSMO SkyMed images to detect land subsidence rates in Shanghai and validated the results by spirit leveling measurements, the results indicate the location of a significant subsidence funnel, the root mean square error of differences of the point targets between two band SAR images is3mm/yr. Luo et al.[2014] extracted land deformation information with L-band ALOS PALSAR data and X-band TerraSAR-X data, the results indicate that the subsidence rate is in the range of-90mm/yr to-10 mm/yr from TerraSAR-X data and -190 to-10mm/yr from PALSAR data, validations of X-band and C-band subsidence results with leveling data and GPS data show good agreement.

 

In this study, two adjacent tracks of TerraSAR-X stripmap images collected from 2010 to 2013, and two adjacent tracks of Envisat ASAR images acquired between 2007 and 2010 were employed to investigate land subsidence in theBeijingregion. Our InSAR results show a maximum subsidence rate of >100 mm/yr in the east part ofBeijing. Comparisons of InSAR-derived subsidence rates between adjacent tracks were performed. A high correlation coefficient was observed with a small RMS difference, indicating the reliability of InSAR retrievals of land subsidence rate maps. GPS derived subsidence rates were also used to compare with InSAR retrievals, and a high consistency was also obtained, which provides additional supporting evidence for the robustness of InSAR retrievals.

 

The spatial and temporal distribution characteristics of land subsidence inBeijingwere analyzed. It is clear that C-band Envisat ASAR provides land subsidence information with a wider spatial scale whilst X-band TerraSAR provides more detailed information with a finer resolution over a smaller coverage. The combination of C-band and X-band data shows obvious advantages for monitoring land subsidence, particularly in urban areas. Based on InSAR derived subsidence rates, a comprehensive model is being carried out to investigate the subsidence mechanisms in theBeijingregion using geological and hydrogeological datasets. 

 

References

Adam N., Eineder M., Schattler B. TerraSAR-X Interferometry Report on a First Assessment. DLR in Proc of IGARSS08.

Berardino P., Fornaro G., Lanari R. A New Algorithm for Surface Deformation Monitoring Based on Small Baseline Differential SAR Interferograms . IEEE Trans. Geosci. Remote Sens. 2002, 40(11): 2375-2383

Chen J. Wu J. Deformation trend extraction based on multi-temporal InSAR inShanghai. Remote Sens. 2013,5,1774-1768

Ferretti A.，Prati C.，Roeca F. Non-linear subsidence rate estimation using permanent scatterers in differential SAR interferometry. IEEE Trans. Geosci. Remote Sens. 2000，38 (5)：2202-2212

Ferretti A.，Prati C.，Roeca F. Permanent Scatterers in SAR interferometry. IEEE Trans. Geosci. Remote Sens. 2001，39 (1)：8-20

Gong H., Zhang Y., Li X. Land subsidence research inBeijingbased on the Permanent Scatterers InSAR technology. Progress in Natural Science. 2009, 19(11):1261-1266

Hooper A., Persistent Scatterer Radar Interferometry for Crustal Deformation Studies and Modeling of Volcanic Deformation.StanfordUniversity, 2006

Jia S.,Wang H. A tentative study of the mechanism of land subsidence inBeijing. Analysis and Research. 2007,2(1):20-26

Lanari R., Mora O. A small Baseline approach for investigating deformation on full resolution differential SAR Interferometry. IEEE Trans. Geosci. Remote Sens. 2004, 42 ( 7 ): 1377 -1386

Li Y., Zhang J., Li Z. Land subsidence inBeijingCityfrom InSAR time series analysis with small baseline subset.Geomatics and Information Science of Wuhan University. 2013,38(11):1374-1377

Li Z., Pasquali P. MERIS atmospheric water vapor correction model for wide swath interferometric synthetic aperture radar. IEEE Geoscience and Remote Sensing Letters. 2012, 9(2): 257-261

Liang F., Sun J. Accumulated crustal deformation and its characteristics inBeijingand Surrounding Regions in 2007-2010 from L-band InSAR. Earthquake. 2013,33(4):43-54

Liu G. Exploration of Subsidence Estimation by Persistent Scatterer InSAR on Time Series of High Resolution TerraSAR-X Images.IEEE journal of selected topics in applied earth observations and remote sensing.2010

Luo Q., Perissin D. L- and X-badn multi-temporal InSAR analysis ofTianjinsubsidence. Remote sensing. 2014, 6:7933-7951

McDonald, R. I., Weber, K., Padowski, J., Flörke, M., Schneider, C., Green, P. A.. Water on an urban planet: Urbanization and the reach of urban water infrastructure. Global Environmental Change, 2014, 27(0), 96-105

Mora O., Mallorqui J. Linear and nonlinear terrain deformation maps from a reduced set of Interferometry SAR images. IEEE Trans. Geosci. Remote Sens. 2003, 41 (10 ): 2243 -2253

Pepe A. Sansosti E. On the generation of ERS/ENVISAT DInSAR time-series via the SBAS technique. IEEE Geosci. And Remote sensing letters. 2005,2(3):265~269

Prati C., Feretti A., Perissin D. Recent advances on surface ground deformation measurement by means of repeated space-borne SAR observations. Journal of Geodynamics. 2010, 49:161-170

WegmullerU., Walter D. Nonuniform ground motion monitoring with TerraSAR-X persistent scaterer Interferometry. IEEE Trans. Geosci. Remote Sens. 2010，48 (2) : 895-904

 

The crustal motion of Great Britain is fundamentally influenced by the process of Glacial Isostatic Adjustment (GIA). This is the mainly vertical response of the solid Earth to the large-scale glaciation and deglaciation of the Earth’s surface. Today’s GIA signal is predominantly influenced by the melting of the Pleistocene British-Irish ice-sheet and also to a lesser extent by the Laurentide and Fennoscandian ice-sheets. The melting of the ice load resulted in uplift of the formerly depressed lithosphere in the glaciated areas, mainly northern Britain with about 1-2 mm yr^-1, accompanied by subsidence in southern Britain [1].

 

Analysing GIA in Scotland is critical for understanding the dynamics of relative sea-level (RSL) change at the coast. RSL is determined by two combined factors: crustal uplift caused by GIA and changes in eustatic sea-levels due to changes in ocean volume. Both contribute to a significant temporal and spatial heterogeneity and variability of the RSL change in Scotland. Against the backdrop of climate change and the global rise of sea-levels caused by the ocean’s thermal expansion and global melt-water influx, the question arises as to what are the modern rate and spatial distribution of GIA. This is especially important, since the process of GIA has generally slowed throughout the Holocene. Also, recent measurements of rising RSL at tide gauges around northern Britain suggest that GIA-related land uplift in Scotland might well be being outpaced by eustatic sea-level change, causing new concerns about the impact on Scottish coasts. This contradicts the belief that sea-level rise in Scotland is “automatically” mitigated by GIA.

 

Applying an Interferometric SAR (InSAR) time-series technique allows determination of  GIA-induced vertical land movement for the past 20 years with a high accuracy and on a broader spatial scale than conventional geodetic techniques. InSAR thus helps to locate the centre of maximum present uplift in Scotland, thereby addressing a current apparent discrepancy between its location derived from GPS measurements on the one hand and GIA modelling results on the other.

SAR data sets covering the land mass of Scotland from different sensors, including ESA ERS-1/2, ESA Envisat, JAXA ALOS PALSAR and possibly DLR TerraSAR-X and ESA Sentinel-1 are being used to establish time-series of surface movements over the past two decades with the Small Baseline InSAR technique [2]. Continuous GPS measurements of the long-term 3D-displacement of the Earth’s surface will not only be employed to mitigate orbital ramps in the InSAR results in this study, but also be used as an independent data set for comparison and validation. Results of this study will eventually inform stakeholders such as environmental agencies, namely Scottish Natural Heritage, for future coastal planning purposes.

 

[1] HANSEN, D. N., TEFERLE, F. N., BINGLEY, R. M. & WILLIAMS, S. D. P. 2012. New estimates of present-day crustal/land motions in the British Isles based on the BIGF Network. In: KENYON, S., PACINO, M. C. & MARTI, U. (eds.) Geodesy for Planet Earth - International Association of Geodesy Symposia 136.

 

[2] BERARDINO, P., FOMARO, G., LANARI, R. & SANSOSTI, E. 2002. A new algorithm for surface deformation monitoring based on small baseline differential SAR Interferograms. In: IEEE Transactions on Geoscience and Remote Sensing, 40, 2375-2383.

In this study, carried out within the EC FP7 MED-SUV project (call FP7 ENV.2012.6.4-2), we present a particular synergy among GPS and SAR techniques, to improve the precision of the current ground deformation monitoring. The DInSAR technique plays an important role in the world of the earth observation techniques used in geodesy to monitor the volcanic areas. Among various perturbations affecting interferograms, atmospheric artifacts are one of the most significant and, probably, the most difficult to identify and reduce. Due to the prominent topography of Mt. Etna and the quite variable weather conditions, the atmospheric heterogeneities become even more pronounced.

For these reason the estimation of Mount Etna atmospheric anomalies using GPS measurements have noticeable importance to calibrate the SAR interferograms and to establish the “effective” ground deformation of the volcanic edifice.

Since 1988, the Istituto Nazionale di Geofisica Vulcanologia, Osservatorio  Etneo (INGV-OE) monitors ground deformations at Mt. Etna. Nowadays, the network geometry consists of 42 permanent stations that provide a dense coverage of most areas of the volcano edifice. The GPS data processing has been carried out by using the GAMIT software, by adopting appropriate processing parameters. A software was developed for deriving the tropospheric tomography starting from the GAMIT output. The code was validated by using synthetic tests which assume different structure of atmospheric anomalies and with random noise about twice severe than the typical errors of the GPS. The results of the tests proved that the tomography software is able to reconstruct the simulated anomalies faithfully. The results of the tomography previously obtained showed clearly important features of the refractivity field of the studied days.

The wet refractivity tomography was applied on experimental data also, these last from DInSAR Sentinel1 IW and GPS measured carried out on Mount Etna during the August 2014.

In order to reduce the known problem of the correction for the antenna pattern, the interferometric process was performed only on one burst of one subset of Sentinel1 IW data. The interferometric results shown a minor effect probably due to the atmospheric artifacts, located on the summit area of Mt. Etna, and the integration with tomographic data allow us to accept or reject this hypothesis. Strategies to calibrate the interferograms from the results of the atmospheric tomography will be discussed.

Radar interferometry (InSAR) method uses radar images for deformation monitoring within large areas. Four ENVISAT stacks of ASAR images acquired from 2003 till 2009 provided by ESA and several TerraSAR-X images acquired from 2011 till 2013 were processed.

The area of interest covers highly built-up area of Prague city. Some subsidences occurred there in the past mainly because of the human activity. There are still some tunnels and roads under construction in the city so there is an assumption of the deformation within examined area. 

First the permanent/persistent scatterers (PS) method was used on ENVISAT images for deformation detection. The processing was performed by IPTA (Interferometric Point Target Analysis) package, which is a part of the GAMMA software. Only appropriate point targets were used for further processing, while the rest of points are omitted. Almost all available images can be used compared to the conventional interferometry method in which the critical baseline requirement must be fulfilled otherwise the interferograms are totally decorrelated.

Each ENVISAT track was processed separately and the results were compared and analysed. Unfortunately there is a lack of images in some tracks and then IPTA method doesn't work very well and the results were unconvincing.

That's why the data were processed also by StaMPS (Stanford Method for Persistent Scatterers) method which is a software package that implements an InSAR persistent scatterers (PS) method. Not the whole scene but only appropriate point targets are used for processing (pixels with high intensity and low intensity variability). This method was developed to work even in terrains devoid of man-made structures and/or undergoing nonsteady deformation. The results from StaMPS and IPTA method for ENVISAT images were compared.

Finally a set of TerraSAR-X images were processed by StaMPS method. The results presented in this paper are just preliminary and will be further analysed, interpreted and also compared (within the bounds of possibility) with the results from ENVISAT images.

In the eastern part of the Czech republic, a new highway construction has started in 2006/2007. The Lysuvky village, near the Frydek-Mistek city, has been divided into two parts by this construction. The newly-built highway leads partially in a tunnel in this area. However, geologically, the area is slightly sloped and with sands underneath, and it was discovered to be sliding down.

Since the start of the construction, the area is monitored by standard geotechnic tools. However, as these supply only pointwise information, InSAR method was used to determine the extent of the landslide area, both in space and in time (whether the landslide was active before the start of the construction).

For processing, ENVISAT data from two neighbouring (descending) tracks, acquired in 2003-2010, were used. Unfortunately, there are a few one-year gaps in the data. There are also very long perpendicular baselines in one track, forcing us to use the Small Baselines technique, and it was not easy to "connect" all the acquired scenes by coherent interferograms in this case. So, in order to connect all the interferograms into a single cloud, lower-quality interferograms had to be used as well.

The highway construction started approximately in the middle of the data acquisition period, causing a non-linear velocity. In addition, the velocity of the deformations is close to the precision limits of InSAR and it is not easy to decide whether the area is really sliding down or not. Therefore, we decided to use time series instead of velocity for interpretation.

A big disadvantage of the Small Baselines technique is its accuracy. It is straightforward to express the standard deviation of deformations estimated for each acquisition time; however, this procedure does not take into consideration the quality of each interferogram. Estimated standard deviation of the velocity can be biased by the linear model, which is not really right. Coherence of each point does not take into account different qualities of the interferograms either.

We therefore look for an accuracy estimation that would be different for each point (as each point has different coherence) and, more importantly, for each acquisition time.

We decided to find a large reference area (with approx. 300 PS points, a suburb of a nearby city), and compute the standard deviation for each scene (considering the area to be stable). The PS points in this area were divided into 5 groups according to their coherence, and these values of standard deviations were used to estimate the accuracy in the area of interest (for PS points of similar coherence).

In addition, two velocity values were computed: one for all the scenes, and one for the second half of the scenes (both with appropriate weights for each scene). Both values are relatively uncontinuous in space (we attribute this to the fact that the estimated deformations are comparable to the accuracy of the method); however, it seems that the velocity estimated for the second half of the acquisition period is a little more continuous.

Processing was performed in the StaMPS software.

One of the main solutions applied to big cities improvement is the construction of large underground infrastructures, in order allow communications when surface space is lacking. Monitoring ground subsidences is therefore one of the main challenges in changing urban environments. 

 

"Madrid Rio" project (2003-2008) is an effort to reclaim the riverfront land and improve the busy M-30 beltway that involved the construction of 7.93 km of tunnels underneath the southern center of Madrid City. This paper presents a remote-sensing approach to monitor ground subsidences induced by tunneling excavation. 

 

Persistent Scatterers Interferometry technique (PSI) was used to estimate subsidence and displacement time series from Synthetic Aperture Radar images, acquired between August 2003 and April 2008 by applying SPN technique to a stack of ENVISAT data.  Remote sensed results were compared to traditional extensometric measures, fitting adequately for selected sectors. Spatial analysis of displacements allowed evaluating impacts of tunneling on surrounding buildings and facilities, highlighting critical areas.  The availability of a spatial distribution of displacements in a time series allowed analyzing longitudinal, cross-sectional and temporal dynamics.

 

Main limitations were the heterogeneous spatial distribution of Persistent Scatterers, the absence of measurement points in work areas, the threshold for velocity detection and low temporal resolution of ENVISAT images. Nevertheless, these limitations of DInSAR for monitoring infrastructures are overcome by actual satellites offering very high spatial and temporal resolutions, being a complementary technique with an exceptional added value and temporal analysis capability.

 

Sentinel 1 data allows now monitoring ground subsidences overcoming traditional problems, thanks to its enhanced terms of revisit, systematic data provision, very large coverage, timeliness and reliability of service. Despite the resolution of the data is not very high compared with other current acquisition modes (spotlight at 0.5 meter resolution), the potential of the data provided by this mission will allow using radar in the construction phases of tunnel works and other infrastructures, complementing traditional control measures.

This work presents the first results of a subsidence detection and monitoring study in an oil shale-mining field in the Ida-Viru County of Interferometric Synthetic Aperture Radar (InSAR) techniques. Ida-Viru County, located in the north-east corner of Estonia, is extensively covered with both subterranean and open-pit oil-shale mines. Mining in this region has been active for nearly 100 years, which means that many of the mines have been inactive for several decades. Collapsing of any of these abandoned subterranean mines is a serious risk to the areas situated above. At least two such collapses have taken place in the previous decade. The aim of this study is to investigate the presence of subsidence in this region to discover any potentially dangerous subterranean mines. Persistent Scatterers Interferometry (PSI) time series analysis was carried out on 13 (out of 53) ERS SAR scenes from track 322 covering years 1995 to 1999. More recent Envisat ASAR data was not used due to excessively sparse coverage in time. The already focused SAR scenes were processed into interferograms using DORIS. From these interferograms persistent scatterer time series were constructed using the Stanford Method for Persistent Scatterers (StaMPS). The study correctly identified slow uplift in the northern coastal towns of Toila and Sillamäe (1–4 mm/yr) which agrees with existing gravimetric measurements. Subsidence due to compaction in the recently-mined areas of open-pit mines was also correctly identified. Within the available dataset it was not possible to reliably measure the vertical ground movements of the areas above subterranean mines, however. These mines tend to be covered with forest or other dense vegetation, which causes the coherence and permanent scatterer density in these areas to be low and the signal-to-noise ratio of the PS's time series to be poor due to phase unwrapping errors. Another limiting factor in the analysis was the low temporal density of available and suitable SAR imagery because of seasonal snow cover and wide differences in effective baselines between acquisitions. As the analysis of Ida-Viru mining region will continue with data received from the recently-launched Sentinel-1, the temporal density and effective baseline issues are expected to be significantly alleviated by Sentinel-1's much shorter revisiting times and tighter orbital tube respectively.

In the Molucca Sea region of northeastern Indonesia, the Sangihe and Halmahera arcs are a unique example of a collision between facing volcanic arcs. Located in eastern Indonesia, which resulted from the complex interaction of four lithospheric plates (Eurasia, Philippine Sea, India, Pacific) and numerous small ones, the Molucca Sea Plate is being subducted simultaneously by both of its western and eastern boundaries. Volcanic arcs have been developed on these margins and provide us a good opportunity to study the different stages of this active arc-arc collision process.

 

The Sangihe arc, the western arc of the collision, is approximately 500 km in length and up to 70 km in width, extending northward from the northeast tip of Sulawesi through Sangihe Island groups to south-central Mindanao, Philippine. Even though the arc contains 25 volcanoes along its length, active volcanoes (12 in total) are constrained to the southern 300 km long part between Awu Volcano on Sangihe Island and Soputan Volcano on north Sulawesi.

 

The four-armed island of Halmahera, on the eastern side of the collision, resembles the western k-shaped Sulawesi in a small scale. It also has a western magnetic arc, which forms the most of the north arm of Halmahera and the islands west of the south arm. The volcanoes within this area (16 in total) form a narrow belt that trends southwestward across the north arm then turns to southward through the offshore islands.

 

In this paper, we perform a systematic survey of the complete Halmahera volcanic arc and the active southern part of Sangihe volcanic arc using Interferometric Synthetic Aperture Radar (InSAR) with Advance Land Observation Satellite (ALOS) data from early 2007 to early 2011. We use a modified version of the Gamma software for raw data focusing and ROI_PAC to generate interferograms, then use our homemade PySAR software for the time series analysis in small baseline approach. Both orbital error and time series DEM error are estimated. With more than 200 scenes of L-band SAR data, our survey reveals the time series deformation activity of 28 volcanoes during these four years, which includes several activity periods of some particular volcanoes.

In this paper we address land subsidence due to excessive groundwater withdrawal in Fars province of south Iran. To retrieve the information on subsidence we utilize Synthetic Aperture Radar (SAR) data from a variety of sources including Envisat, ALOS and TerraSAR-X satellites. The SAR data are processed using the technique of Small BAseline Subset (SBAS) to assess ground deformation in space and time. The subsidence maps are analysed together with information obtained from head declines and subsurface geology to assess rheological properties of the aquifer system in the area.

 

The western Galápagos Islands (Ecuador) host some of the most rapidly deforming and frequently erupting volcanoes on Earth. Determining the amount of magma coming from the mantle and supplied to the crustal reservoirs beneath each volcano becomes fundamental to any attempt to understand their past evolution or to anticipate their future behaviour. In this study we use space-geodetic measurements of the surface displacements at the six active volcanoes on the westernmost islands of Isabela and Fernandina to estimate volumes and rates of magma supply to the volcanoes and, by summing each contribution, to the entire archipelago during 1992-2011.

Surface displacements are measured using interferometric synthetic aperture radar (InSAR) data from five different satellites: the European Space Agency’s ERS-1, ERS-2 and Envisat satellites, the Canadian Space Agency’s Radarsat-1 satellite, and the Japanese Aerospace Exploration Agency’s ALOS-1 satellite. The complete dataset consists of more than 500 SAR acquisitions from 17 tracks in both ascending and descending orbital passes of the satellites. Global positioning system (GPS) data from a local network at Sierra Negra, the most actively deforming volcano in the archipelago, complement the InSAR observations.

The storage system of each deforming volcano is characterized in terms of its location and geometry using analytical models and through the non-linear inversion of the InSAR data. We observe that surface displacements at all the six deforming volcanoes can be well explained by volume changes in crustal reservoirs with depths ranging between 1 and 6 km below the surface. Time series of vertical displacements for each volcano are then formed by combining InSAR data from ascending and descending satellite-passes, and used to track volume changes within the magma reservoirs during the eighteen-year time period.

We calculate that a cumulative volume of 0.350±0.100 km³ was supplied at an average rate of ~0.02±0.005 km³ yr^-1 during 1992-2011 to the Galápagos volcanoes and that more than half of it was directed towards Sierra Negra. The rate at which magma is supplied, however, largely varied through time in correlation with eruptive and intrusive activity at the volcanoes. A significant reduction in magma supply to the archipelago is observed between 2000 and late 2003, a period during which eruptions did not occur. Successively, a surge in supply is revealed by rapid inflation a Sierra Negra that led to an eruption in 2005. Very high rates of supply are also observed after the eruption at Sierra Negra during 2005-2006 and after a subvolcanic intrusion at Fernandina in 2007, suggesting possible pressure unbalances between the volcanoes’ reservoirs and the deep roots of their plumbing systems. On the other hand, eruptions and intrusions during the past two decades have only occurred at the three volcanoes showing the highest rates of magma supply (Sierra Negra, Fernandina and Cerro Azul), suggesting a possible feedback between the two processes.

Our results provide the first archipelago-wide magma supply rate estimate in the Galápagos on a time-scale of interest to humans and constitute the most detailed geodetic study of the Galápagos Archipelago. The outcome, based on almost twenty years of InSAR observations, demonstrates the value of this technique in regional-scale studies of active volcanism.

Mount Etna is an active stratovolcano located on the eastern coast of Sicily, Italy. For studying ground deformation at Mt Etna we collected ascending and descending RADARSAT-2 Standard-3 (S3) data spanning January 2009 – June 2013. The advanced multidimensional small baseline subset method (MSBAS, Samsonov and d’Oreye, 2011) was employed to produce horizontal and vertical time series of ground deformation. Vertical deformation map revealed localized subsidence and horizontal deformation map revealed broad east-west motion at the fault located north of the volcano with rates <2 cm/year. The nonlinear inversion method of Camacho et al. (2011) was used for modeling of pressure or/and mass sources with a free geometry. The method assumes simple homogenous elastic conditions, and models the size and location of pressure/mass sources in a general model, fitting the entire geodetic data set within some regularity conditions. The approach (Camacho et al., 2011; Samsonov et al., 2014) works as a growth process that constructs a very general geometrical configuration. A general framework for the geometry of anomalous bodies is achieved by partitioning the subsurface volume into a dense 3-D grid of point sources. The geometrical description of the extended sources is developed by aggregation of point sources in a 3-D context, allowing for a very general description of the free geometry for the source bodies. For simplicity, it is assumed that anomalous pressure changes are nearly homogeneous within the sources. The inversion process fills cells with the prescribed pressure value (and/or density anomaly), giving rise to the aggregation structure. The nonlinear search is based on an exploratory approach.

 

REFERENCES

Camacho, A. G., P. J. González, J. Fernández, and G. Berrino (2011), Simultaneous inversion of surface deformation and gravity changes by means of extended bodies with a free geometry: Application to deforming calderas, J. Geophys. Res., 116, B10401, doi:10.1029/2010JB008165.

Samsonov S. and d’Oreye N., 2012. Multidimensional time series analysis of ground deformation from multiple InSAR data sets applied to Virunga Volcanic Province, Geophysical Journal International, 191 (3), 1095-1108, DOI: 10.1111/j.1365-246X.2012.05669.x.

Samsonov, S. V., K. F. Tiampo, A. G. Camacho, J. Fernández, and P. J. González, 2014. Spatiotemporal analysis and interpretation of 1993–2013 ground deformation at Campi Flegrei, Italy, observed by advanced DInSAR. Geophys. Res. Lett., 41, 6101-6108, doi: 10.1002/2014GL060595.

InSAR monitoring of landslides along strategic transportation and energy corridors, are instrumental to security and development planning. With current available spaceborne SAR systems and upcoming constellations (Sentinel-1b, RADARSAT Constellation Mission (RCM), and proposed Tandem-L), routine monitoring of infrastructures by InSAR is becoming mainstream practice.  

 

In preparation to RCM, various unstable sites triggered by different processes (coastal erosion/tidal, permafrost activity, hydrology/snow melt, and anthropogenic) are being monitored by CCRS and GSC. The following case study consists about a complex landslide affecting a coastal railway in Gaspe, Quebec, Canada. The terrain consists of Paleozoic sedimentary rocks sloping toward the Baie des Chaleurs at an angle between 15 and 25 degrees while the local slope is about 60^o. This section of the railway has been impacted by slope movements in the last several decades and required continuous maintenance and multiple repairs. Previous field studies carried out in the 1990’s by researchers from Université Laval revealed a complex network of large open fissures beneath and in the vicinity of the railroad. Monitoring data obtained in the mid 1990’s showed that some rock blocks and fissures are characterised by significant displacement. In the summer of 1998 (July 23^rd), a rock failure occurred and the railway was close for several days. In order to better understand the behavior of the terrain and to develop a mitigation plan, a geotechnical study combining in situ measurement and InSAR techniques has been put in place.

 

Two categories of in situ instruments are distributed along the railway: 1) which are extensometer rods, inclinometer boreholes and total station targets which require manual surveys and; 2) shape acceleration arrays (SAA), tiltmeter, piezometers, and crackmeters which are linked to a computer which relays to an University server by FTP twice a day. Since the forested section above the railway strongly reduces the InSAR coherence at C and X bands, the 13 trihedrals and ten dihedrals where installed in 2008 and 2009, for InSAR coherent point target analyses (figure 1). Three trihedrals and two dihedrals are installed outside the unstable zone and serve as reference targets for phase calibration and orbital baseline refinement. Remaining target devices are distributed in the forested zone and along the railway. Seven of them are set up for SAR ascending pass acquisitions while 16 are for descending, both under right looking configuration.

 

Since the fall of 2008, RADARSAT-2 Spotlight data have been acquired at 27^o and 44^o incident angles, descending pass, every 24 days and since the fall of 2009, RADARSAT-2 Spotlight, ascending pass, are acquired at 32^o. Complementary, TerraSAR-X seasonal acquisitions were taken between 2009 and 2013 in both ascending and descending configurations, with an incident angle of 32^o. InSAR displacement estimates show that the landslide is characterized by a more complex series of active blocs than previously estimated by traditional in situ instruments. They reveal important displacement rates in a zone initially considered stable in the 1990’s surveys. InSAR analyses from ascending and descending orbit configurations have enabled a better understanding of the spatial distribution of the displacement in 2D and in 3D while in situ instruments provide only one dimension component (except SAA) and some requiring a manual survey. Coherent point target analyses have been assessed over corner reflectors located in high biomass zones (above the railway) while stacking techniques have been applied over the cliff below the railway.  InSAR point target analyses done with TerraSAR-X show similar trends as with RADARSAT-2, where different zones are moving at different rates (from 5 to 40 mm/year). Furthermore, the trihedral installed on the retaining wall shows an acceleration of its displacement rate in October 2010, which has raised a flag to the railroad manager for closer monitoring. The11 days repeat orbit cycle and the higher resolution capabilities, of TerraSAR-X outperformed RADARSAT-2 over the unvegetated areas (rail and lower portion of the site) with the stacking approach. These valuable InSAR measurements are currently being integrated with other in-situ measurements to model the kinematics of this landslide.

 

Keywords: RADARSAT-2, TerraSAR-X, InSAR, landslides, transportation corridors.

Dams are multi-purpose projects having significant social and economic impact, such as flood control, irrigation, water supply and electricity production. Monitoring a dam plays an essential role in its management and operation, especially for unsafe conditions or problems that require appropriate corrective measures at the early stages. Possible failure can cause great disasters with enormous social, economic and ecological costs.  During the last decades, remote sensing techniques (e.g. laser scanner, differential interferometry, etc.) came to complement the methods for measuring the displacement of infrastructure with in-situ instruments as GPS, etc. Among these techniques, the Differential Interferometry SAR (DInSAR) is used to measure the micro-movements with millimeter precision.

The aim of this paper is the observation of surface deformation of Mornos Dam region and its link with physical parameters, exploiting multitemporal-interferometry methodology on satellite SAR acquisitions (Figure 1). Specifically, we have detected and analyzed the deformation due to landslides at North and South abutment as well as on the Mornos dam itself. Its behavior has been related to possible sources of deformation such as seismic events that occur in the broader region and water level changes of Mornos artificial lake. The data used in this study were 29 ascending and   39 descending acquisitions of ASAR/ENVISAT covering the period of 2003-2010. For the production of deformation maps, the technique of Deferential Interferometry (DInSAR) SBAS (Small Baseline Subset) and for the time series analysis, the SVD algorithm (Singular Value Decomposition) were applied. The maximum variation of the deformation on the dam for the period 2003-2010 is about 4 cm along the Line of sight (LOS). Although the observed deformation appears to be due to changes in water level following a particular pattern, there are discontinuous over time which coincide with specific seismic events. Due to the availability of both ascending and descending data, the east-west and up-down components focusing only in the Mornos dam are presented. It is remarkable that processing with data set of AMI/ERS acquisitions covering the period of 1992-2000 was made and the results shown similar to the ASAR.

We would like to acknowledge ESA for the ASAR/ENVISAT data through cat-1 6287 and Athens Water Supply and Sewerage Company (EYDAP SA) for the water level of the Mornos artificial lake data. 

Surface deformation processes related to volcanic activity occur on various temporal scales (seconds to years) affecting areas on various spatial scales (meters to kilometers). Over the past decades, Synthetic Aperture Radar Interferometry (InSAR) has been widely used for detecting volcano related surface deformation. At the same time, satellite systems develop towards higher spatial and temporal coverage, as well as improved spatial resolutions, enabling the application of InSAR to a wide range of volcanic deformation processes. While most deformation studies at volcanoes still focus on volcano-wide surface deformation in order to assess the overall state of eruptive activity, we combine high-resolution TerraSAR-X SpotLight imagery with a precise digital elevation model from Terrestrial Laser Scanner (TLS) data in order to study very local deformation patterns related to the morphology and evolution of volcanic craters. Here we present new results achieved at the nested summit craters of Láscar Volcano, Chile.

TLS data were collected over two days in December 2013 using a long-range RIEGL instrument. We calculate a Digital Elevation Model (DEM) of the active eastern edifice of Láscar Volcano, featuring a spatial resolution of more than 1 m from over 15 million data points. This high resolution DEM allows identifying and mapping local structures. Using the SBAS algorithm, we then analyze TerraSAR-X SpotLight interferograms of the Volcano’s summit area at 2 m resolution over the period between September 2012 and May 2014. We detect two separate regions of local subsidence occuring within the volcano's summit craters. Another very local area in between two craters might be slowly uplifting. We consider these movements to be related to the general geomorphologic setting and structural architecture of the nested summit craters. Our results imply that local and small scale structures identified first by TLS are indeed showing small scale displacements in high-resolution InSAR observations. These results could not be achieved with any other method or lower spatial resolutions, demonstrating the great prospects for the continued use of high-precision remote sensing datasets in order to investigate inaccessible and ever-changing environments, like volcanic summit areas and craters.

On May 12, 2008, a catastrophic earthquake (Wenchuan Earthquake) measuring 8.0 on the Richter scale struck Sichuan Province in southwestern China and a series of aftershocks occurred after the earthquake. In Wenchuan earthquake affected area, 316 aftershocks measuring over 4.0 Richter scale took place by July. 26, 2010. More than 40 among these aftershocks measured 5.0, and up to a maximum of 6.4 Richter scale. A large number of houses collapsed, traffic system and communication equipment damaged to varying degrees. The powerful earthquake caused a major series of ecological and environmental problems; including impacts on major types of land use and land cover changes (LUCC), soil erosion, mountain hazards, all of which occurred were present at many sites following the earthquake. As of Sep. 25, 2012, the great earthquake caused more than 69000 death and 370000 injured, direct economic losses stood at 800 billion Yuan.

The Yuleishan area of Longmenshan, lies in the northwest Chengdu Plain, located between 103°40' 104°55' eastern longitude and 30°8' 31°26' northern latitude covering an area of 450 square km. Yuleishan is a mountainous region of steep topography and intertwining channels. Elevation ranges from 870m to 4810 m with a north-south decline. The area north of Pengzhou City, near the towns of Longmenshan and Xiaoyudong experienced the most severe earthquake damage in the greater Longmenshan area.

Geographically, the region lies in the transition between the Tibetan Plateau and the Chengdu Plains, thus, both geology and geomorphology conditions are complex. In the LMSS, the climate is typically humid sub-tropical monsoon, with mild winters, hot summers, and a long frost-free period with plentiful rainfall. The annual rainfall is 1225.7 mm, with 88 % occurring between May and October. The yearly average temperature and yearly average sunlight duration are 15.7°C and 1131 hours, respectively.

The main river, the Jianjiang, is more than 18 km long forming a pinnate network pattern. Much of the standing vegetation in the primary branches of the Jianjiang was completely destroyed, especially in Yinchanggou and Baishuihe. The natural vegetation is classified as subtropical evergreen broad-leaf forest.

This region is remarkably rich in biodiversity and important scenic areas, such as Yinchanggou National Geological Park, Yinchanggou 4A Grade Scenery, Baishuihe National Forest Park, and Baishuihe National Nature Reserve.

Longmenshan Scenic area has been facing with debris-flow hazards and landslides more frequently in recent years during rain seasons. Many settlements are threatened by high disaster risks in this area. The area of Mountain Hazards increased dramatically in the study area after Wenchuan earthquake. There were five reported debris flow events triggered by monsoon storms in the study area between June and September in 2008. Similar group of debris flows and landslides occurred in this area in summer 2009 (17th, July), 2010 (18th, August), 2011(2nd, July, 2012(17th, August) and 2013(11th, July). Some of these debris flow events demolished settlements, damaged the Yin-bai Highway, blocked traffic, buried farmland, and further endangered human life and property. It is clear that there is considerable increased and continuing risk of debris flows and landslides, especially during the rainy monsoon season.

At a large scale, the total area of mass wasting in the post-earthquake period sharply increased from 302.98 km² to 389.28 km² Compared with pre-earthquake, an increase of 86.30 km² (accounting for 19.2% of the study area). The greatest change in soil erosion occurred between 1000 m and 4000 m, while negligible change occurred below 1000 m and above 4000 m. The previous denuded area was 17.46 km² but increases to 83.26 km² after the earthquake. The denuded area increased 65.81 km², mostly attributed to the Wenchuan Earthquake.

However, there was a significant interaction between degree of slope and percentage of Japanese cedar, the mainly planted trees in the vegetation cover. Lines would be parallel if there were no interaction. Instead, the frequency of slopes sliding does not uniformly depend on the slope, or on the amount of Japanese cedar present in the vegetation cover, but on the effect that different levels of Japanese cedar have on slopes of varying degrees. On slopes with gradients lower than 40 degrees, the probability of sliding is low for both high and low percentages of Japanese cedar. On medium slopes, the probability of sliding is high for high levels of Japanese cedar and low for low levels of Japanese cedar. On the steepest slopes, the probability of sliding is extremely high for high percentages of Japanese cedar but still relatively low for low percentages of Japanese cedar.

Because of our limited field data and the complex topography, as well as frequent secondary disasters, it was not possible to acquire sufficient data to statistically estimate the potential risks caused by the earthquake over this area.

Until now, there has been little data to quantify information about environmental changes at the regional scale following the earthquake because of limited sources of data and complicated problems in accessing mountain sites to record the impact of the seismic event. However, documenting environment changes is necessary for appropriate decision making to reduce future secondary earthquake disasters.

The aim of this work concerns the implementation of multitemporal SAR interferometry in order to assess slope stability in the Yuleishan of Longmenshan. Specifically by implementing a “hybrid INSAR” methodology using elements of conventional DINSAR, multi-baseline interferometry approaches. To obtain the least-squares solution for the phase time-series we applied Singular Value Decomposition (SVD) algorithm. ASAR/ENVISAT* SLC dataset of 34 descending scenes (Track: 290, Frame: 2976) covering the period 2003-2010. From the above dataset a subset of 25 were used covering the post-earthquake period 2008-2010. Rough topography and dense vegetation restricted high coherence mainly over existing urban areas. Our results reveal negative values of deformation with a maximum rate of deformation locally up to -20cm along the Line Of Sight. From interferometric results mountainous areas are excluded due to SAR geometry effects and low coherence. Finally the post seismic sequence events, following the main shock, of 4.0-5.5Mw observed in single differential interferograms could contribute to the deformation measurements of the area.

 *Data provided by the European Space Agency. The research is carried out within the frame of Greece-China bilateral projects supported by GSRT of Greece.

Volcanic eruptions have significant economic and societal impacts, yet the majority of the world’s volcanoes have no ground-based monitoring. Satellite imagery covers volcanoes globally throughout their eruptive cycle and has the potential to provide an alternative, globally applicable source of information. Although appropriate InSAR data have not yet been analyzed over all of the world’s volcanoes, the number of known deforming volcanoes has more than quadrupled since 1997 (e.g, >200 today vs. 44 in Dvorak & Dzurisin, 1997) and ~500 volcanoes have been observed as part of systematic, regional studies providing a multi-decadal archive suitable for probabilistic analysis linking deformation with eruption.

Here we report on the progress of the Global Volcano Model’s Deformation Task Force, and report the results of probabilistic analysis of the data already created (Biggs et al, 2014). Volcanoes that erupted during the satellite observation window are 5.6 times more likely to have recorded deformation than not (the positive likelihood ratio), meaning this dataset provides ‘strong’ evidential worth. However, the positive predictive value (PPV) of 0.44 (the proportion of volcanoes that deformed during the observation window, that also erupted) indicates that deformation alone, while worthy of concern, should not be considered a strong diagnostic of imminent eruption. Moreover, there is a much greater negative predictive value (NPV) of 0.94, i.e. the proportion of ‘non-deformed’ volcanoes that did not erupt in the recording period. Although much emphasis is placed on the ability of volcanologists to predict eruptions, often the ability to make a valid negative deduction can be very valuable for hazard and risk decisions. We further show that the link between deformation and eruption is influenced by tectonic, petrological and volcanological factors.  Due to the publication bias towards periods of deformation, we are not yet able to quantify the causal or temporal relationships between the extent, rates, timing, amount and duration of deformation and subsequent eruption using InSAR alone.The launches of Sentinel-1 and rapid repeat observations from constellations such as CosmoSkyMed will greatly increase temporal resolution and coverage, enabling more subtle distinctions to be incorporated into pattern analysis. Our probabilistic analysis thus provides a framework by which satellite observations can be used to inform hazard decisions and strategic development for the future.

 

 

 

A dam is a barrier that impounds surface water or underground streams. The reservoirs created by dams not only suppress floods but also provide water for various needs including agricultural irrigation, human consumption, industrial use, aquaculture, navigability and power generation. Earth-fill dams are constructed as a simple embankment of well compacted earth. A zoned-earth dam has distinct parts or zones of dissimilar material, typically a locally plentiful shell with a watertight clay core.

Deformation is unavoidable for any dam and thus, Monitoring is an essential component of dam after construction and during operation. In general, conventional methods for evaluating dam deformation are able to acquire high-precision measurements in discrete points; however, high financial and labor costs and time consuming are unavoidable. On the other hand, Interferometric Synthetic Aperture Radar (InSAR) helps us to derive surface deformation over a wide area by analyzing phase difference between two or more SAR images.

In this research, we present preliminary results of deformation analysis at Masjed-Soleyman earth-fill dam on Karoon River in southwest of Iran using high-resolution SAR images acquired by TerraSAR-X satellite. Masjed-Soleyman dam is a large earth-fill dam with a large size central clay core that has been constructed in 2000. Dam has a vertical clay core and its height is 177 m. 13 SAR images in a descending orbit and 6 SAR images in an ascending orbit were used to assess dam stability. All images were acquired in spotlight mode and were processed by the Small BAseline Subset (SBAS) technique using high-resolution TanDEM-X Digital Elevation Model (DEM) to derive deformation maps. The results are compared to independent measurements obtained from classical geodetic surveys for quality assessment.

SkyGeo offers continuous deformation monitoring based on a number of InSAR techniques. We show two examples where we provide updated deformation measurements quickly after new images are acquired. These measurements help our customers in the oil and gas industry as well as managers of infrastructure assets do their jobs efficiently.

The first example is based on 11-day updates of vertical and east-west movement on an oil field. The highly nonlinear movement caused by exploitation of the field and subsequent injection of water is monitored using high-resolution X-band (TerraSAR-X) data. This helps guiding the injection strategy and preventing failures in boreholes.

The second example is a continuous long-term large-scale monitoring effort over a large part of the Netherlands. In particular, we help underground infrastructure managers follow the slow deformation due to weak soil and the changes therein. A new PS-InSAR technique based on Kalman filtering is used to provide highly-accurate deformation measurements around pipelines. This technique allows for consistent updates and efficient adjustment of recent measurements, even in the case of nonlinear deformation.

Over the Netherlands, excellent C-band radar imagery has been available almost nonstop since 1992. The ERS and Envisat missions were followed up by nationwide RadarSat-2 acquisitions to bridge the period before Sentinel becomes operational. The currently ongoing RadarSat-2 acquisitions, and recently started Sentinel measurements provide continuity in the medium-resolution coverage.

Between the nineties and today, deformation measurements from space have found numerous applications in aspects of environmental, safety, and economic management of infrastructure on and below the Earth's surface. Almost all these applications, however, have been shown to benefit tremendously from switching to higher resolutions. Higher resolution allows for finer detail, reliability, wide coverage, and ease of interpretation. This opens up a wide range of applications. A great example here is provided for two sluice complexes in Germany, where initial analyses were made with ERS/Envisat, then high-resolution studies were made with TerraSar-X, and now a transition to spotlight is on its way.

New developments at MDA have made it possible to acquire similar nation-wide coverage at five times higher resolution. These new observations allow for the continuation of the medium-resolution activities, but also allow for high resolution analyses. In this contribution, we report on feasibility studies of switching to higher resolution within an ongoing medium-resolution stack. We show how the existing stack enables a 'warm start' of the high-resolution analysis, and how the high-resolution imagery can be used to provide continuity with older analyses.

Traditionally, slope instabilities are monitored using ground-based methods such as tachometry, precise levelling and global positioning system (GPS) surveying. However, these techniques are time-consuming and cost-intensive and limited regarding spatial and temporal coverage. Further, only point measurements are provided, minimizing the possibility of adequately depicting the dynamics of complex natural phenomena. The increasing amount of nowadays available Earth observation (EO) data has led to the evolution of novel techniques for the mapping and monitoring of unstable slopes. Especially SAR interferometry has proven to be powerful for detecting and monitoring small-scale surface deformations associated with slope instabilities with a centimeter to millimeter accuracy.

In contrast to well-established methods for slope instability investigation by means of SAR interferometry (e.g. Persistant Scatterer Interferometry), the present study aims at semi-automatically identifying and delineating unstable slopes and active landslides based on single-pair SAR interferograms and coherence images using object-based image analysis (OBIA). OBIA makes use of information derived from image objects, which are created by merging adjacent pixels with similar spectral characteristics. OBIA provides a framework for examining features defined by spectral, spatial, textural and contextual properties by combining image processing and GIS functionalities. Objects are derived through image segmentation and are - unlike single pixels - enriched by a range of features/properties (e.g. mean values, minimum and maximum values, mean ratios, standard deviation). For the proposed study several SAR data stacks, particularly TerraSAR-X data, are employed to generate SAR interferograms and coherence images using Next ESA SAR Toolbox (NEST). The derived products are orthorectified and selected filters (e.g. Goldstein, Convolution) are applied. Subsequently, data is integrated in eCognition (Trimble) software for semi-automated identification and delineation of unstable slopes and landslides, i.e. areas which were subject to change between data acquisitions. Image segmentation is conducted to generate appropriate image objects, which serve as basis for the identification of unstable slopes and landslides. The obtained polygons are finally exported as vector data. In order to assess the accuracy of the results, the identified objects are compared to manually digitized reference polygons or delimitations derived from optical remote sensing data. Test sites are located in Taiwan, which is highly susceptible to mass movements due to its particular geologic and climatic settings, and in the alpine terrain of Salzburg, Austria and Upper Bavaria, southern Germany, both of which are frequently affected by landslides, especially after heavy rainfall events.

This study shows that SAR interferometric products can be applied for the semi-automated delineation of unstable slopes and landslides in an object-based image analysis domain. The developed approach may be useful for supporting the investigation of active landslides, but also for the update of existing landslide inventories. The object-based delineation of unstable slopes may further provide relevant information (e.g. spatial extent of landslide activity) for risk assessment and landslide forecasting.

The Piton de la Fournaise, in La Reunion island, is one of the most active basaltic shield volcanoes of the world. Thanks to the initial effort of the Observatoire de Physique du Globe de Clermont-Ferrand, France, Piton de la Fournaise is also one of the few volcanoes in the world that is monitored with InSAR data on a regular basis. About forty eruptions between 1998 and present have already been imaged and analyzed using various radar satellite data (e.g., Radarsat-2, ASAR-Envisat, ALOS-PALSAR, TerraSAR-X, TanDEM-X and Cosmo-SkyMed). In this work, we present maps of the co-eruptive displacements related to the October 2010, the December 2010 and the June 2014 eruptions. These maps were produced using whichever of the following satellite data is/are available: Radarsat-2 (RST-2), Cosmo-SkyMed (CSK), TerraSAR-X (TSX) or TanDEM-X (TDX). The results gave us a clear idea about the mechanism of the eruption (i.e., propagating dyke towards the surface). Then, by computing the coherence of the co-eruptive interferograms, we have also been able to map the extent of the lava flows and to estimate their areas.

In addition, we have derived accurate thickness map of each lava flows using two alternative techniques. The first one involves a large database of interferograms produced from monostatic CSK images. The second one is based on the use of few bistatic TSX-TDX interferograms. Results show that CSK database does not only deliver thickness, volume and effusion rate values but it can also provide displacement rate information after the lava flows were emplaced. However, we emphasize that for this case, accuracy is highly dependent on the number of interferograms used. Evidently, a large database such as the one that we have used will require a long time to be built and may not be very compatible with near-real time monitoring. The bistatic dataset, on the other hand, can only provide thickness, volume and effusion rate measurements but with accuracy even with just two interferograms. Furthermore, even for thin flow deposits, the bistatic dataset constantly showed exceptional results.

For cross-validation purposes, we have compared our results with MODIS-derived volume and effusion rate values as well as with available aerial photos and optical images acquired by the volcano observatory of Piton de la Fournaise.

In Belgium, the federal collection of built heritage is visited every year by millions of people. This built heritage is of exceptional cultural and economic importance for the country and its protection is a priority at federal and regional levels. Due to heavy industrial and urban development, cultural heritage buildings suffer from physical, mechanical, chemical, and biochemical pathologies along their history. In addition to alter the appearance of the monument these factors inherent to the building material affect the structural behaviour of the construction and its stability. Moreover, external human factors such as groundwater pumping, underground galleries, excavations contribute to disturb the stability of the buildings. Frequently, collections of archives such as coal mines, galleries maps help to delimit areas subject to ground instabilities. On the other hand underground excavations of phosphate or clays are known but their exact localization is unknown due to their usually artisanal state. For instance, subsidence generates ground settlement due to a growth of soil effective stresses following a reduction in piezometric level. This phenomenon is usually not spatially uniform due to soil properties and spatial variations of deformable soil thickness and piezometric levels, causing differential settlements and distortions that affect the buildings foundations. The measurement of the progression and distribution of these settlements is necessary in order to adopt the appropriate corrective or mitigation measures on the subsidence caused by aquifer or subsoil exploitation

 

Declercq P.-Y. (2012a) thanks to the TERRAFRIMA I and recently PANGEO projects has shown using the Persistent Scatterer method that the city of Brussels is on the opposite uplifting due to a historical diminution of the groundwater pumping. Water demanding industries such as breweries have left the city centre for the periphery in relation with economic pressure and land use management. Additionally, Belgium has a long history of coal mining activities which stopped 30 years ago. Calembert (1955), created a map showing the extension of the influence of the underground activities in the coal basin of Liège. During the exploitation time the ground movement was corresponding to a subsidence but after the closure of the mines in the 1970’s the groundwater raised and provoked an uplift (Devleeschouwer, 2008). The annual average velocities recorded by Persistent Scatterer analysis was between 1- 2 mm/year over a period of 15 years. Declercq (2012b), Cuenca (2012) have respectively shown a similar behaviour in the Belgian and Dutch coal area of Limburg.

 

The project GEPATAR (GEotechnical and Patrimonial Archives Toolbox for ARchitectural conservation in Belgium) aims the fusion between the archives of the Geological Survey of Belgium (GSB) and the Royal Institute of Cultural Heritage (RICH) for better management of the Belgian federal patrimony by estimating the stability of the monuments by radar interferometry over the entire country. The project is financed by BRAIN (Belgian Research Action through Interdisciplinary Networks) under the Belgian Science Policy umbrella. In the first stage of the project, a PSI processing chain software will be developed by Centre Spatial de Liege (CSL) for the exploitation of the ERS and ENVISAT SAR data available at GSB. By the creation of PS-InSAR processing models, it will be possible to map and analyse with accuracy of millimetres, the ground movement trend at local and regional levels. A toolbox will be created by the Signal and Image Centre of the Royal Military Academy (RMA) in GIS environment allowing the integration of data from the RICH archive with inputs from RBINS at three scales: national, regional and local. ‘National’ refers in this study to the Belgian country, ‘regional’ refers to the region in risk of ground movement and ‘local’ is the patrimony building. The integration of topographical, cadastral, land cover, geological, geomorphological and ground stability with historical, structural engineering and architectural data will allow the risk analysis for Belgian cultural heritage. At National and regional levels, GEPATAR will provide to the staff involved in defining conservation strategies at RICH the possibility to prioritize the conservation actions. At local level (i.e. building level) GEPATAR will provide all the necessary data set to assess the risk for a patrimony building. Evaluation of GEPATAR toolbox will be done using identified test cases. The selected monuments are located at areas where intensive urbanism and post-industrial activities have occurred. A ground movement risk model will be developed for each building for which a large set of environmental, geographical, historical and architectural data sets needs to be collected. The information will be combined using the GEPATAR modules for further structural engineering analysis, assessing the strength of the building structure by the Catholic University of Leuven (KUL).

 

 

 

Calembert, L. (1955). Géologie, Mines et Aménagement régional du Bassin industriel liégeois. Revue universelle des mines. 9è série, 11(12), 645-669.

Cuenca, M. C., Hooper, A. J., & Hanssen, R. F. (2012). Surface deformation induced by water influx in the abandoned coal mines in Limburg, the Netherlands observed by satellite radar interferometry. Journal of Applied Geophysics. (88), 1-11.

Declercq, P.-Y., Pirard, E. and Devleeschouwer X. (2012a). Brussels 20 years of Uplifting monitored by radar interferometry (PSI). ). In: Abstract book, 7th EUropean Congress on REgional GEOscientific Cartography and Information Systems (EUREGEO), Bologna.

Devleeschouwer, X., Declercq, P. Y., Flamion, B., Brixko, J., Timmermans, A., & Vanneste, J. (2008). Uplift revealed by radar interferometry around Liège (Belgium): a relation with rising mining groundwater. In Proceedings of the Post-Mining Symposium (pp. 6-8).

Declercq, P.-Y., Devleeschouwer X , Brassinnes S. and Pirard, E. (2012b). Preliminary results of an unexpected uplift situated in a former coal mining region (Campine Basin – Belgium) revealed by radar interferometry. In: Abstract book, 1st Workshop on Temporal Analysis of Satellite Images, Mykonos, Greece, May 2012

 

Highway bridge in the southern part of Prague, Czech republic, is the longest bridge in the Czech republic and it is architectonicly unique in Europe. It was built above the valley with confluence of two big rivers, Vltava and Berounka, in an area of frequent floods. On top, there is a five-lane highway with an important, circle-shape crossroad, and both ends of the bridge end up in a tunnel.The bridge was constructed in 2006-2010.

The total length of the bridge exceeds 2200 meters however only small parts of it are above water; it has two parts constructed by different companies. Maximum height of the bridge is 65 meters, and the bridge is bending.  The smaller part of the bridge, with a length exceeding 200 meters, is supported by oblique pillars. However, during construction, steel was partially replaced by concrete in comparison to the plan, and first cracks arised just after opening of the bridge in September 2010, and are progressively getting bigger.

We try to monitor this bridge using the Interferometric Point Target Analysis (IPTA) technique, with approximately 20 TerraSAR-X scenes, acquired in the Staring Spotlight mode with a resolution of approximately 0.25 meters during June - December 2014 (11 days interval). Temperature is also monitored, on a close small bridge approx. 500 m away from this large one, and part of the estimated deformations is attributed to temperature changes which exceed 30 degrees.

Unfortunately, the orientation of the bridge with regard to the satellite tracks is not ideal, the bridge is approximately perpendicular to the descending track, and even in the ascending scenes, only the top of the bridge can be seen, with no pillars. In addition, in the processed results, there are only three thin “lines” of processed pixels, however the bridge is at least 15 m wide.

This abstract was written in order to present the results of my master´s thesis (Mai.2014 – Dec.2014). The thesis was inspired by the work of Bert Kampes’ “Radar Interferometry – Persistent Scatterer Technique” (2006). The study presents results obtained by data acquired from 1992 to 2000 by European Remote Sensing satellites ERS-1 and ERS-2.

The aim was to gain knowledge about the technique of radar interferometry. This technique will be demonstrated by the reconstruction of Bert Kampes’ practical example of Berlin. The focus has been on the theoretical surface deformation around a gas reservoir in the midwest of Berlin (Germany). The main challenge for radar interferometry in this region is the predominance of forest and lakes. The phonological effect has a significant influence on the coherence of long term deformation monitoring, hence on the results.

In order to solve this problem, the study was realized with the small baseline method. The main advantage of this method – instead of the persistent scatterer technique – is the limitation of the baseline decorrelation phenomena.

Synthetic Aperture Radar (SAR) interferometry has found many applications as a technique to measure terrain displacements of areas affected by mass movements. Furthermore, it has proven to be useful also a means to map areas interested by a flood event. For this kind of application, both SAR amplitude and interferometric coherence are used to discriminate areas covered by water at the time of SAR acquisitions and estimate the whole extension of the area interested by the flood. The aim of this work is to use of time series of Cosmo-Sky-Med SAR images to study an event characterized by both flood and mass movement phenomena.

The study area is located in the city of Chosica Lurigancho district, province and department of Lima, on the left bank of the Rimac River, to 38 km east of Lima. It has an altitude of 850 meters. The main access is through the Central Highway, there are two alternative paths that cross the Rimac river to the area of study, or even directly from Chosica income from the Bridge Los Angeles. Part of the fan and the broken channel is occupied by recreational club Regattas and farmland at the bottom.

The climate corresponds to the coastal strip of Peru, desert temperate and humid, characterized by low rainfall throughout the year except from January to March. The minimum average total rainfall per year is 18 mm; but the normal daily maximum rainfall of 22 mm.

The area presents a much defined mainstream, a basin of very large and open reception on two sides at his bedside; the main channel has a SE-NW direction. The slopes or hillsides have slopes between 20 ° to 35 °. The main channel of the creek has a length of 4890 m in the floodway and an average slope of 10°.

The area is not covered by huge vegetation and is characterized by the presence of houses, roads and other infrastructures.

On Thursday, April 5, 2012 at 17:30 a heavy rain over three hours, focused on areas Chosica, Ricardo Palma Avenue (Chaclacayo) and triggered debris flow, mud and rocks that exceeded the capacity of the channel in lower section (channelized section). The debris flow had a great intensity. This occurred between 27 and 42 kilometers from the Central Highway in Lima. It caused destruction of housing, water and sewage networks, road blocks by the impact of huge rocks and mud that flooded roads, streets and avenues.

The destructive effects occurred mostly in the left margin; while on the right bank where mitigation works are to flow, the damage was minor. The debris flow had a precedent in 2009.

In this work we estimate the extension of the area interested by the flood and to measure the terrain displacement caused by debris flow phenomena triggered by the event. A time series of single-look-complex SAR images acquired by the Cosmo-Sky-Med satellite over the monitored area is interferometrically processed and results presented in terms of maps of flooded areas and of terrain displacement of the areas affected by the mass movements.

Landslides are recognized as natural hazards having a major social–economic impact and represent a significant risk for citizens as well as infrastructures. During the last years, a significant increase of landslide occurrences has been recorded globally as a result of increased urbanization and development, continued deforestation and severe meteorological events. In Greece, this increase exceeds the 20% of the totally recorded cases during the last fifteen years. In this study, preliminary results of the “Landslide Vulnerability Model – LAVMO” project concerning the monitoring of a number of case studies will be presented.

 

LAVMO is a 3-year project aiming at creating a persistently updated electronic platform which will help to monitor, predict and assess risks, through a decision making model, related with landslides. Within this framework, three landsliding areas were selected to be investigated as case studies in the Gulf of Corinth area (Figure 1). The selected sites are characterized as the most important “active” landslides in the region of Achaia Prefecture. These are the landsliding zones of Platanos and Panagopoula as well as the landslides of Karua-Souli. They are under the influence of geodynamic processes, while the lithology, stratigraphy and morphology combined with seismic and tectonic activity strongly influence the slope instabilities.

 

In the framework of the LAVMO project, a multitude of sensors were considered and different methodologies were followed in order to assess the best solutions. All the available observations will be fused in order to provide a comprehensive information representation. In the initial phase of the project a multitemporal interferometric study was performed exploiting archived ASAR/ENVISAT acquisitions to assess the sensor detecting and measuring capabilities for the specific landslides. In parallel, very high resolution new SAR acquisitions of TERRASAR-X, RADARSAT-2 and COSMO-SkyMed, over the two test sites (Panagopoula and Platanos) were used to produce mutitemporal interferometry deformation maps. For the third site (Karua-Souli), archived high resolution SAR data from RADARSAT-2 were also used for producing interferometric maps.

Complementary to the InSAR data, cross-correlation techniques were exploited for observations in the panchromatic channel of very high resolution multispectral sensors (QuickBird primarily and World View secondarily).

 

Throughout the duration of the project, continuous in-situ measurement are being conducted in the three landslide areas. More specifically, measurements from inclinometer into boreholes in and around the landsliding bodies, sampled every two meters and up to a depth of 50 meters from the surface, are being performed in conjunction with GPS campaigns. Future sub-surface measurements using a Geo-Radar are expected to contribute to the study in the next phase. Satellite and ground measurements are being acquired throughout the duration of the project.

 

AKNOWLEDGMENTS

We would like to thank ESA, CSA and DLR for providing SAR data through the science projects ESA TPM CSA SOAR-EI 5171, DLR MTH20111 & MTH1686, ESA TPM ASI-EGEOS C1F.14971 and ESA cat-1 6287. The project is funded by European Union (European Social Fund - ESF) and National Strategic Reference Framework (NSRF) ‘Thales’ project.

This work is focused in the infrastructure monitoring of areas which had experienced an increase of urban areas and therefore, also an increase of the exploitation of natural resources. Persistent Scatters Interferometry (PS-InSAR) and Small Baselines (SBAS) approaches are applied to three study areas for which large datasets of SAR images are available in ascending and descending modes to finally deployed deformation maps of different buildings and infrastructures.

Valencia, Rome and South Sardinia are areas which had experienced an increase of the exploitation of the natural resources parallelly to the urban expansion. Moreover, Rome is a very special case, where Cultural Heritage filled the city and its surroundings, being necessary a tool for monitoring the stability of the different sites. This work wants to analyse the potential deformation that had occurred in these areas during the period 1992 to 2010, by applying Persistent Scatters Interferometry to the ESA ERS SAR and Envisat ASAR IMS datasets.

Slope deformations are the most important geohazards in Slovakia which annually cause an extensive economic damage of significant influence and limit the rational use of land and in rare cases also threaten human lives. In Slovakia, about 22 000 slope deformations have been registered so far, covering an area of almost 2 600 km² and corresponding to 5.25% of the territory of the Slovak Republic. This is conditioned by the complex geological and tectonic setting of the Slovak territory, which, in conjunction with other important elements, creates favourable conditions for their development. In the most affected areas of the Flysch region landslides may threaten up to 40% of the area and in the peripheral areas of Neovolcanic region it can be up to 60% of the territory. The slopes within Neogene and Paleogene depressions are also significantly affected. Moreover, in the current period the activation of relatively large amounts of landslides have been witnessed. Since 2010, 639 new slope failures have been registered and their activation was driven mainly by the climatic anomalies, such as extraordinary rainfalls and melting snow cover. Many of these landslides currently represent a direct threat to the lives, health and property of the residents in the affected areas. This fact has initiated their inclusion in the monitoring system built in the scope of the project “Partial Monitoring System - Geological factors“. The stability condition in these areas is studied by regular measurements based on the observations of groundwater level depth changes, geodetic and inclinometric measurements and regular evaluation of climatic factors. For the areas of high socio-economic importance where the large deformations are occurring, up-to-now monitoring activities has currently been extended to account for the results from multi-temporal InSAR techniques implemented in SARRPOZ software. The selected site of Nizna Mysla is densely populated municipality that rest exclusively upon active landslides. The assessment of the physical activity of the landslide phenomena becomes crucial after extraordinary rainfalls in 2010 which has induced a lot of slope failures in some regions of Slovakia, cumulating into most disastrous landslide in Nizna Mysla village on 4^th of June, 2010. The local infrastructure was damaged and several houses have been destroyed or became uninhabitable. From the time series of the cumulative displacement obtained by the processing of 64 ENVISAT ASAR scenes from ascending track No. 329 (32) and descending track No. 408 (32) acquired within years 2003-2010 and using standard PSInSAR model expanded for the non-linear component, changes on some structures were observable since 2009 (See Appendix 1, Figure 1). More images from ERS and ENVISAT satellites are evaluated since then as well as in-situ measurements for completeness, verification and comparison reasons. The applied methodology appears to be promising for the complementary updating of existing landslide inventory maps aimed at proper urban planning with benefits of satellite radar interferometry, concerning cost-effective monitoring with wide area coverage, no need for human or special equipment presence in the area of interest, backdating capability and accuracy comparable to the conventional monitoring techniques. With already ongoing SENTINEL-1 mission and free accessible near-real time data archive, the future work will be focused on the setting of possible precursors for the slope failures in order to estimate landslide risks.

Interferometric SAR (InSAR) allows measurement of millimiter-level surface displacements. This is accomplished by pixel-by-pixel phase comparison (interferometry) of two or more images separated in time, combined with accurate models of satellite orbit and other geophysical processes affecting the transmitted and reflected microwave signals. InSAR displacement estimate using (SBAS) are used to observe land subsidence with noise in the area above the subway construction of Budapest.

Given our current understanding of the pysics there is no doubt that tunnel construction is associated with time-dependent subsidance, and that this subsidence can be detected with space-based SAR measurements. The key questions relate to whether current generation commercial sensors have sufficient measurment accuracy (several millimiters or less), a function of phase coherence and the number of scenes available for analysis; the spatial resolution of the sensor (a few meters); the temporal resoution of the sensor; wheter the requried coverage is avaiable, and the anaylitical techniques used.

We intend to employ techniques in Data mining of time series databases  to isolate pixels in the interfergromas containing timesires that can be grouped in a certain type of time-dependent transient response subsidence related to the under ground tunnel excavation. This is then, followed by a principal component anaylsys to determine the direction of the tunnel via sorting of the indexes of the selected pixels. Ultimately, Kalman filtering is used with a  linear model to perform a spatial filtering in two dimensions (Range and Azimuth) of the foot print of the tunnel on the SAR image.

The final product of our system includes, the foot print of the tunnel, the length, width, and duration of construction (computed from the time siries along the time domain). High resolution spotlight data (HS) from TerraSAR is intended to be used for our project, with a revisit frequency of 11 days.

Konya Closed Basin, having an area of approximately 55 000 km², and covering the provinces Konya, Karaman, Aksaray, and Niğde, is one of the important centers of Turkey in terms of agricultural and industrial production. The water needed for the population,  ecological system, and agricultural and industrial production should be gained from limited surface waters like Beyşehir Lake which capacity is not sufficient for the whole water demand. This is the reason why groundwater sources are often used for water supply. The significant decrease of the water level of the wells, distributed in the basin, and operated by State Hydraulic Works, has been observed since 1960. The extent of this decrease is more significant in the last decade. In some wells, the decreases of 1 m per year have been observed.

            The decrease of groundwater caused by the rapid increase in population, global warming, over-use of water resources, and other influences is one of the most important problems of the world. The problems coming from the decrease of groundwater are not limited to loss of water resources. Depending on the decreasing water volume, a deformation and compaction are observed on aquifer systems that hold the groundwater. This effect triggers the subsidence of the land above the aquifer systems. The subsidence rate can reach tens of cm or dm level. According to the USGS database, 80% of the recorded land subsidence cases of more than 100 worldwide records are related to the groundwater decline [1]. The agricultural irrigation regions that use groundwater resources, metropolis, and mining areas are under threat as a result of land subsidence effect. Finally, the land subsidence damages structures such as roads, water canals, drainage systems, pipelines, and buildings in cities.

            The aim of this study is to determine and to monitor the areas affected by land subsidence caused by the decrease of groundwater, by using geodetic methods in Konya Closed Basin. The study area is entirely on the Anatolian plate and covers a limited area of Konya closed basin. It is assumed that the tectonic activity effect is far below the values of vertical deformations due to the land subsidence. The GPS and InSAR techniques which complement each other perfectly will be used to deduce spatial deformations caused by the land subsidence.

            The land subsidence activity in the area of Küçükköy situated at the Konya Closed Basin (KCB) of the Central Anatolia Region in Turkey has been studied using the Stanford Method for Persistent Scatterers (StaMPS) InSAR technique. Significant movements of land subsidence ranging from mm to cm per year were detected with ENVISAT ASAR data between 2003 and 2010 above the city center. Results are also compared with GNSS measurements. Estimated deformation rate of KCKY GPS station is approximately 2 cm/yr. A strong correlation of about 90 % between subsidence values and groundwater usage over the area of interest has been determined.

            This paper briefly presents the results of StaMPS PS and SB InSAR processing of Envisat ASAR data for a case study in Küçükköy district (boundaries of KCB). The magnitudes of subsidence detected by these techniques are close to GNSS results. Farming of the fields for agricultural purposes makes difficult detecting surface deformations from the interferograms but the results show meaningful land subsidence effects of several cm per year caused by groundwater withdrawal in study area. Finally, the paper discusses applicability of Sentinel-1 and other SAR data (especially L-bad SAR) for further monitoring of the area of interest.

References:

Land Subsidence, The USGS Water Science School, USGS, 2014 [online: http://water.usgs.gov/edu/earthgwlandsubside.html]

 

Keywords: Konya Closed Basin (KCB), PS-InSAR, StaMPS, land subsidence, groundwater.

Interferometric Synthetic Aperture Radar (InSAR) is a satellite-based remote sensing technique that has the ability to observe ground deformation at a sub centimeter-scale. This technique can be used to monitor ground movement for rectangular areas as large as 100 kilometers on a side. Knowledge of topography, geology, trends and mechanics of existing ground movement is required for successful interpretation of InSAR data. The objective of this study was to evaluate evidence of seasonal (short-term) and long-term ground uplift and subsidence in the vicinity of a natural gas storage facility in the Coastal Plain province of Mississippi using InSAR data. The uplift and subsidence evaluation was performed using archived InSAR data from different time periods. Analysis of the historical ground surface change was carried out using inSAR data from three satellites (ERS-1, ENVISAT and ALOS). The data from ERS-1 and ENVISAT were provided by the European Space Agency (ESA). The earliest data from the satellite ERS-1 was used to interpret change between July 1992 to September 1992 for one month and two months period. InSAR data from ALOS was used to interpret change between January 2008 and July 2010 for 27 months and 30 months period. InSAR data from ENVISAT from October 2005 to March 2009 was used to interpret change varying from a period of 7 months (short-term) to a period of 40 months (long-term).

 

The interpreted InSAR values show clearly that interpretable subsidence was not observed for the gas storage facility area during any period analyzed. Slight subsidence/uplift of 1 to 2 cm was observed in short-term periods that may be due to cyclic loading effects, possibly related to gas injection and withdrawal. In the long term, however, it is clear that very little subsidence occurred within the vicinity of the gas storage facility. Our study also revealed that subsidence of 2.5 centimeters was interpreted in a two-month period in 1992 in an area north of the study site. Similarly, subsidence of 6.5 centimeters was observed between 2008 and 2010 in an area east of the study site.  The interpretation of subsidence in the areas adjacent to the study site demonstrate the validity of the InSAR calculations and potential of InSAR analysis to detect uplift and subsidence in the study area. One of the advantages of InSAR analysis is the ability for subsidence to be detected in areas other than study area because the data covers a larger area. This finding from the InSAR analysis sometimes may prove to be vital for identifying future subsidence or issues in virgin areas. Our study also proved that the InSAR technology may be used to detect and monitor ground subsidence at similar gas storage facilities throughout the world.

As one of the most important natural wetland ecosystems in the world, peatlands cover 3% (some 4 million km²) of the Earth’s land area and store up to 550 Gt of carbon, equivalent to one-third of global soil carbon [Hooijer et al., 2006]. However, peatlands can degrade on a large scale, particularly in the tropical areas with land use changes [Charman et al., 2008; Couwenberg, 2011]. Through both legal and illegal logging, and conversion to agricultural use over the period of 1985~2006, about 12.1 M ha of peatland was deforested and drained in Southeast Asia, of which 1.5 M ha were tropical peat swamp forests in Central Kalimantan, Indonesia [Hooijer et al., 2006]. Drainage causes peat oxidation and releases carbon to the atmosphere as CO₂, which results in peat subsidence [Wösten et al., 1997]. It is estimated that an additional drainage depth of 10 cm results in 0.9 cm peat subsidence and about 9 t CO₂ ha^-1yr^-1 emissions [Couwenberg et al., 2010]. Another severe consequence of drainage is that the peat surface becomes dry and the peat is susceptible to fire during the dry season (Generally from June to October). Fires are most severe in very long dry seasons, such as in the 1997/98 El Niño Southern Oscillation (ENSO) event. It is reported that 0.12-0.15 Gt of carbon was emitted due to peat combustion from Central Kalimantan, Indonesia during ENSO dry season of 1997, and the total carbon loss in Indonesia is between 0.81 and 2.57 Gt, equivalent to 13-40% of the mean annual global carbon emissions from fossil fuels [Page et al., 2002]. Therefore, restoration of peat hydrology and conservation of tropical peatlands is critical to prevent peat oxidation and mitigate excess CO₂ emission [Wösten et al., 2008; J. Jaenicke et al., 2011; Sasha et al., 2011]. From 2004 to 2008, hydrological restoration commenced in Central Kalimantan by building large dams [CKPP, 2008; J. Jaenicke et al., 2011]. Our objective is to investigate the restoration effects after hydrological restoration by detecting surface height change, because a blocking dam can be regarded as successful if it raises the groundwater and prevent peat subsidence continuously[Wösten et al., 2008; Julia Jaenicke et al., 2010].

Using field-based surveying to monitoring changes in the past peatland surface height and over the large areas typical of drained and restored, is challenging such that measurments are more likely to describe a small area and be only a snapshot in time. Upscaling and understanding the rate of change in surface height trough time may be overcome using emote sensing approaches.

Here we present data on the change in peatland surface height in Central Kalimantan, Indonesia, detected using the Interferometry Synthetic Aperture Radar (InSAR) Small BAseline Set (SBAS) approach[Berardino et al., 2002; Li et al., 2009]. Using data from December 2006 to September 2010, we have generated a map of the rate of change of mean height, and time series of change in drained peatland. To do this we used two independent ALOS L-band tracks SAR images, as there is a lack of ground data for validation, correlation in output provides confidence the results are representative. Our analysis to date shows that in areas with restroation: 1) the restoration is affected by the construction time and density of dams - the earlier the dam is constructed, the more significant the restoration is; 2) high-density dams have positive effects on restoration in a short time. And in areas without dam construction and restoration work: 1) the rate of change in surface height (decrease) can vary; 2) the decrease in surface height can be up to -7.65 cm/year; 3) the largest decrease in surface height observed was 25 cm. Also we consider these peat subside results in the context of the loss of regional and global C storage, our result shows that the study area has an annual CO₂ emission rate of 24.36 t/ha/yr.

This study has demonstrated that InSAR is a powerful tool in investigating the restoration effects, and can also be used to indirectly assess the relative magnitude and spatial pattern of damages of peatland caused by drainage and fires; the latter provides key information for guiding future restoration activities. InSAR not only offers a great advantage through regional coverage compared to field-based observation of individual monitoring point, but InSAR provides a practical way to monitor peatland height changes continuously and globally, with archived JERS/ALOS-1 dataset and the launch of ALOS-2.

References:

Wösten, J. H. M., A. B. Ismail, and A. L. M. van Wijk (1997), Peat subsidence and its practical implications: a case study in Malaysia, Geoderma, 78(1-2), 25-36.

Berardino, P., G. Fornaro, R. Lanari, and E. Sansosti (2002), A new algorithm for surface deformation monitoring based on small baseline differential SAR interferograms, Geoscience and Remote Sensing, IEEE Transactions on, 40(11), 2375-2383.

Page, S., F. Siegert, J. O. Rieley, H.-D. V. Boehm, A. Jaya, and S. Limin (2002), The amount of carbon released from peat and forest fires in Indonesia during 1997, Nature, 420(6911), 61-65.

Hooijer, A., M. Silvius, H. Wösten, and S. Page (2006), PEAT-CO2,Assessment of CO2 emissions from drained peatlands in SE Asia. Delft Hydraulics report Q3943.

Charman, D. J., H. Joosten, J. Laine, D. Lee, T. Minayeva, S. Opdam, F. Parish, M. Silvius, and A. Sirin (2008), Assessment on Peatlands, Biodiversity and Climate Change: Main Report, Global Enviroment Centre, Kuala Lumpur & Wetlands International, Wageningen.

CKPP (2008), Provisional Report of the Central Kalimantan Peatland Project. Novermber 2008Rep., 70 pp, CKPP Consortium, Palangka Raya, Indonesia.

Wösten, J. H. M., E. Clymans, S. E. Page, J. O. Rieley, and S. H. Limin (2008), Peat–water interrelationships in a tropical peatland ecosystem in Southeast Asia, CATENA, 73.

Li, Z., E. J. Fielding, and P. Cross (2009), Integration of InSAR Time-Series Analysis and Water-Vapor Correction for Mapping Postseismic Motion After the 2003 Bam (Iran) Earthquake, Geoscience and Remote Sensing, IEEE Transactions on, 47(9), 3220-3230.

Couwenberg, J., R. Dommain, and H. Joosten (2010), Greenhouse gas fluxes from tropical peatlands in south-east Asia, Global Change Biology, 16(6), 1715-1732.

Jaenicke, J., H. Wösten, A. Budiman, and F. Siegert (2010), Planning hydrological restoration of peatlands in Indonesia to mitigate carbon dioxide emissions, Mitig Adapt Strateg Glob Change, 15(3), 223-239.

Couwenberg, J. (2011), Greenhouse gas emissions from managed peat soils: is the IPCC reporting guidance realistic?, Mires and Peat, 8(2), 1-10.

Jaenicke, J., S. Englhart, and F. Siegert (2011), Monitoring the effect of restoration measures in Indonesian peatlands by radar satellite imagery, Journal of Environmental Management, 92(3), 630-638.

Sasha, A., et al. (2011), Opportunities and Challenges for Ecological Restoration within REDD+, Restoration Ecology, 19(6), 683-689.

SAR Interferometric (InSAR) techniques are widely applied for analyzing geophysical and tectonic deformations and for improving modeling techniques, with the benefits of surface deformation measurements covering areas from several to a hundred kilometers in size with a satisfactory accuracy (theoretically in the millimeter range). Nevertheless the accuracy and availability of deformation measurements using InSAR techniques is limited due to decorrelation effects, atmospheric disturbances, the SAR side-looking geometry (which causes layover and shadowing). In this paper, we will present advanced InSAR techniques combined with cross-correlation measurements in order to retrieve non-linear deformation signals from active volcanoes using high resolution TerraSAR-X mission images. An experimental TanDEM-X Raw DEM with a resolution of approximately 6 meters has been applied in order to compensate the topographic phase. A volcanic test site which is currently highly active - El Hierro Island - has been selected in order to demonstrate the developed advanced techniques.

Due to the rugged geological and geomorphologic environments, Zhouqu County in Gansu Province of China is very vulnerable to the landslide hazards. In this paper, the StaMPS-SBAS algorithm is employed to identify and measure the slow movements of the landslides over the Zhouqu County. A total of 16 L-band ALOS/PALSAR images acquired during January 2007 and June 2010 are used to accomplish this purpose. In particular, we propose a novel phase ramp correction in this study to improve the performance of the StaMPS-SBAS method in the suppression of orbital errors and atmospheric artifacts. This proposed method employs the multi-look unwrapped interferograms to model and correct the phase ramps induced by orbital errors and atmospheric artifacts in the single-look wrapped interferograms. The phase ramp model is constructed relevant to SAR acquisitions instead of interferograms to better minimize the effects of phase noises in the phase ramp correction. The slow-moving and continuously creeping landslides in four local regions, i.e., Luojiayu-Sanyanyu, Suoertou, Xieliupo and Nanshanqiaotou, are identified and mapped, which positions and scales coincide well with those from the previous field investigations. The results show that the landslides include segments with different movement rates and the maximum is up to 70 mm/yr. The slopes in the Luojiayu-Sanyanyu valley have been found to be quite instable before the 2010 great mudslide. Deformation exceeding 30 mm/yr has been observed in the Suoertou and Xieliupo slopes. About 30 mm/yr deformation has been detected in the Nanshanqiaotou slope, as a prelude of the 2010 large landslide. Time series of landslide movements were found to be highly correlated to the local precipitation with around 4-5 month time lags between the maximum deformation area and the peak of the precipitation. Such time lags for Zhouqu landslides are the combination result of the abundant loose rock and sediment characterized by soft-hard lithologic groups, strong weathering and structural fissure development as well as the badly-distributed precipitation. In addition, an interesting phenomenon was observed that the landslide movements occurred in June 2008 were much larger than that in other periods. This is expected to be imputed to the Wenchuan earthquake occurred only about 300 km away from Zhouqu area in May 12, 2008.

Magma plumbing systems at active volcanoes are often not well understood. The combined use of time-dependent microgravity (also referred to as dynamic or 4D microgravity) and ground deformation measurements provide a useful way to probe shallow magmatic processes and can potentially help in forecasting future eruptions. In this study, we review current procedures on the joint use of microgravity and deformation at active volcanoes, with emphasis on hazard mitigation aspects.

Time-dependent microgravity is a geophysical method that consists of repeated measurements of the gravitational attraction of the Earth at specific stations either continuously or at the time of survey campaigns. After correcting the gravity measurements for all the known components that contribute to the measured signal (i.e. earth-tides, instrument drift, ocean-loading, water table variations, free-air effect, etc.), residual gravity changes provide information on mass flow beneath the surface. When time-dependent microgravity is applied to volcanoes, the residual gravity changes can be attributed to magmatic activity. Analytical models can then be used to estimate the location and the magnitude of the mass variations responsible for the gravity changes.

 Surface deformation measurements, such as those obtained through time-series analysis of interferometric aperture radar (InSAR) data, can instead be interpreted in terms of subsurface changes in pressure and, by inference, volume changes. InSAR data can be used to constrain surface displacements along the radar line-of-sight and, by combining measurements from independent ascending and descending orbital tracks, the vertical and horizontal components of the surface displacement can be estimated. Similarly to the microgravity data, surface displacement measurements can be inverted using analytical models to infer the location and the volume change within the source of deformation.

Finally, by comparing the mass change from the gravity measurements and the volume change from the surface displacement measurements spanning a given time interval, the density of the fluid responsible for such changes can be inferred. If the inferred density is close to that of gas for example, vesiculation processes can be detected with this approach. Vesiculation and intrusion of fresh magma (volume changes) within a magmatic reservoir are two common mechanisms capable of triggering volcanic eruptions. Being able to detect these processes before other precursory activity takes place (e.g. an increase in the frequency of earthquakes beneath the active volcano, the occurrence of volcanic tremor, an increase of magmatic degassing at the surface of the volcano, etc.) may therefore improve preparedness and hazard mitigation.

Several studies coupling deformation data with time-dependent microgravity measurements have been carried out during the last few decades at different active volcanoes (e.g. Mt Etna, Kilauea, Long-Valley, Campi Flegrei, Askja, Krafla) and proved the success of this approach.

An extensive comparison of these previous studies will be presented here to not only highlight the benefits of the approach but also identify shortcomings and limitations of the joint use of microgravity and ground deformation measurements.

After 40 years of absence of eruptions in the Canary Islands, in July 2011 volcanic unrest started in El Hierro, the youngest of the seven Archipelago islands. Thousands of earthquakes, which reached a maximum local magnitude (M_L) of 4.3, migrated from north to south over the island in three months (1).They were accompanied by maximum deformation of 6 centimeters in the vertical direction several days before the onset of the submarine eruption, on October 2011. Deeper seismicity and deformations continued during the whole eruptive process, until March 2012.

The GPS network on the island, despite of playing an essential role in the process, did not allow to measure displacements in some areas of El Hierro, especially prior to the beginning of the eruption and during the first days of it. To reinforce the spatial coverage of the displacement measurements, a dataset of ascending and descending fine-resolution TerraSAR-X images were acquired and processed. We were able to observe centimeter decreases in line of line of sight (LOS) in the southwest part of El Hierro where no GPS data were available. These changes were associated to the most seismically active area, where later the eruption took place.

Due to the complex atmospheric regime of the archipelago, large tropospheric signals were found in the interferograms obscuring the displacement signals. We have applied several tropospheric correction methods, including a simple phase to height relation and a recently developed power-law correction method (2). Unfortunately both methods were not able to represent the complex and turbulent tropospheric signals in all our interferograms. For this reason, we ran a fine spatial resolution weather model (WRF). Using this independent correction method we were capable of reducing the troposphere signals, and found after correction of the InSAR data a good correlation with GPS estimates of the tectonic signal.

By combining the temporal sampling of the GPS and the spatial coverage of the corrected InSAR data a comprehensive picture of the spatial-temporal deformation during the pre- and syn-eruptive episodes in El Hierro emerges. The study case of El Hierro 2011-2012 eruption highlights the importance of combining GPS and InSAR data for volcano deformation monitoring.

(1)    López C. et al, "Monitoring the volcanic unrest of El Hierro (Canary Islands) before the onset of the 2011–2012 submarine eruption", Geophysical Research Letters, VOL. 39, L13303, doi:10.1029/2012GL051846, 2012.

(2)    Bekaert D.P.S, Hooper A.J., Wright T.J., “A spatially-variable power-law tropospheric correction technique for InSAR data”, Journal of Geophysical Research, in review.

The goal of this study is to evaluate new Staring Spotlight mode capabilities to monitor the mining activities impacts on the environment in order to ensure an effective management and to prevent possible natural and technological hazards. The societal and environmental impacts are huge such as: the topographic alteration, changes in the soil structure and vegetation coverage, influence on the underground water resources and on the rain water draining regime and air pollution. Rosia Jiu opencast test site is part of Motru-Rovinari coal basin which is located in southwest of Romania, in the Getic Piedmont, stretches between Jiu Valley (est) to Motru Valley (west). The area is affected by subsidence phenomena caused by the closing of the hollows remained from the underground exploitation of lignite and by altering of the hidro-geological conditions, due to the applying of a forced and high intensity dewatering of the aquifer system within the area (Braghina, 2011). In this view, the test site Rosia Jiu opencast has a particular specification: underground rock excavation equipment, human settlements located on two sided at distances of 100-150 m and configuration and negative rates of three limits of the opencast. In the study, a methodology based on deformation maps will be design for monitoring of the elastic deformation, early warning stage and detection of the risk occurrence. It will be obtained more precise deformation models of rock strata that could improve mining operations.

Understanding the dynamics of landslides and predicting their occurrence is a common theme in the study of natural hazards. This paper demonstrates the monitoring of mass-movement caused by Europe’s largest coastal landslide complex, Black Ven, on the UNESCO designated Jurassic Coast World Heritage Site (JCWHS) in the south of the United Kingdom, using a coupled InSAR, airborne laser scanning (ALS) and Differential Global Navigation Satellite System (D-GNSS) approach.

Mass-movement on the JCWHS represents a “coastal conflict” between the delivery of fossils to the beach, by landsliding and mud flows etc, which are a draw for professional fossil hunters and tourists alike, and the need to protect the property and infrastructure on which visitation depends. These important fossil sites and the classic coastal geomorphologic features which feed them are also the main reasons for UNESCO World Heritage Site designation. Reliable and frequent monitoring of the site is therefore essential for providing insights into the dynamics of the system, especially the acceleration and/or migration of displacement, and any subsequent management decisions.

Interferometric SAR (InSAR) techniques have been applied to time series of ENVISAT ASAR (C-band) and ALOS PALSAR (L-band) to detect and quantify displacement. Both conventional InSAR and offset-tracking have been applied according to local ground conditions. These same techniques will also be applied to Sentinel-1, with the Sentinel-1A & 1B constellation also representing a considerable reduction of temporal baselines for C-band orbital SAR to just 12 days. It is anticipated this will improve monitoring of ephemeral mass-movement events.

Repeat airborne laser scanning (ALS, also known as airborne LiDAR) overflights by the Environment Agency (England & Wales) have provided a time series of Digital Surface Models (DSM). These have fulfilled two functions, namely 1) monitoring changes in surface volume over time; and 2) providing contemporary data for the removal of topographic phase in InSAR analysis, thereby improving accuracy. Repeat D-GNSS surveys measurements, using markers placed on the landslide complex, provide reference data for InSAR observations and additional data where InSAR is not possible.

I will present synthetic tests and analysis of archive interferograms to demonstrate the utility of source separation methods for identifying independent and interconnected volcano deformation.  

There have been several observations of eruptions at one volcano that are accompanied by unrest or eruption at its neighbours (e.g. Klein et al., 1982; Hildreth & Fierstein, 2000; Miklius & Cervelli, 2003; Wright et al., 2012).  However, it is not clear how reliable historical or geological observations are (Brown et al., 2014), or in fact how common eruption clustering is (Bebbington &. Lei, 1996). Coupled activity may be the result of connected zones of melt in the crust (Gudmundsson & Andrew, 2007), interacting stress fields (Albino & Sigmundsson, 2014) or of a common response to distant stress changes (Manga & Brodsky, 2006).  Linked deformation signals potentially provide evidence of otherwise unobservable processes. Objective tests of the independence of deformation at neighbouring volcanoes and within calderas will allow the identification of coupled activity and potentially the identification of the mechanisms responsible.

Blind source separation methods are used routinely in medical data analysis (e.g. EEG or fMRI) and are very successful in both signal separation and feature extraction.   Independent Component Analysis is the most widely used method for performing blind source separation, and produces a linear representation of data so that components are statistically independent.  This means that it is appropriate for the extraction of low magnitude signals, even where noise is high, without any a priori assumption beyond the independence of the components. Here, I present the results of synthetic tests of such techniques to analyse volcanic deformatio signals.

Albino & Sigmundsson, (2014). JGRS, 119 2964-2975; Bebbington & Lei, (1996). JVGR, 74, 101-110; Brown, et al., (2014). JAV, 3, 5; Gudmundsson & Andrew, (2007). GRL, 34(10); Hildreth & Fierstein. (2000). GSA, 112.10, 1594-1620; Klein, (1982). JGVR, 12(1), 1-35; Manga & Brodsky. (2006). AR, 34, 1, 263-291; Miklius & Cervelli (2003). Nature, 421, 229; Wright et al., (2012). NatGeo, 5, 242–250

Since surface subsidence threaten urban infrastructures, monitoring the surface subsidence in urban areas is very critical. Nowadays, Interferometry helps scientists to monitor this phenomenon. Conventional and PSI methods are used for InSAR analysis. StaMPS method is one of the PSI methods used for monitoring the deformation of the earth surface. This method can not measure the rate of the change in the case of high gradient subsidence accurately. The aim of this paper is presenting an analysis of the performance of StaMPS PSI techniques for high gradient deformation monitoring using TerraSAR-X data. we are able to measure the high deformation rate that was not already possible by radar images with large revisit cycles. The case study for this project is located on the southwest of Tehran, capital city of Iran. 11 TerraSAR-X images acquired from Tehran city spanning between 2011.02.07 and 2012.16.05 have been used to monitor the land subsidence induced by over-exploitation of groundwater. Displacement velocity map was estimated using the StaMPS PSI time series analysis results. The conventional SBAS method is carried out on this stack of TerraSAR-X data to have a comprehensive cross-validation between the results obtained from SBAS and PSI methods. The second goal of the project is comparing the ability of moderate and high resolution data in estimating the deformation rate using PSI method. An available dataset of ENVISAT ASAR in the study area from 2003 to 2008 are processed using StaMPS method to compare results of two stacks of data. Finally, the obtained results are validated with available leveling measurements. According to the results, TerraSAR-X data significantly improved the performance of StaMPS method in estimating high-rate deformation in comparison with ENVISAT ASAR data. Moreover, results of TerraSAR-X data using StaMPS PSI method are closed to the results of SBAS method and leveling data. The reason of this improvement is high spatial and temporal sampling in TerraSAR-X data. These characteristic of TerraSAR-X data improved the phase unwrapping process. 

Persistent Scatterer Interferometric SAR (PSInSAR) is an advanced differential interferometric technique that utilizes stable radar targets to map the deformation. As Mumbai city comprises of many natural and artificial reflectors that work as stable radar targets, in this study PSInSAR technique is used to map land deformation in the city using L and C band data. For generating the land deformation map, 15 images of Envisat and ERS-1/2 covering the period of 14 years, spanning from 1992 to 2005 acquired over Mumbai through European Space Agency (ESA) Cat-1 project and 12 images of ALOS PALSAR covering the period of four years from 2007 to 2010 were processed and displacement maps were generated for C and L band respectively. The deformation results obtained by utilizing the ALOS PALSAR L-band data are also compared with C-band deformation results. For comparing the results, PS points occurring in both the displacement maps were selected and time series deformation was studied and compared. Qualitative analysis of displacement map is carried out and it is found that the built-up area in Mumbai city is undergoing subsidence. The subsidence is divided into high, medium and low subsidence area due to the non-availability of field data for validation. For assigning the cause of the subsidence, ground water level map and geology map of the study area are also studied and correlated with the deformation results. A methodology for the identification of the endangered building has also been developed. For the first time, building subsidence study for Mumbai city was done using PSInSAR technique. It was observed that not all buildings in the study region is undergoing subsidence; in some area uplift is also observed. Ascertaining the causes of the subsidence and uplift of the buildings require additional building parameters like construction year, building material, foundation of the building etcetera.

Among the many benefits of satellite based remote sensing is providing global coverage for various volcano-monitoring techniques.  Satellite-based Interferometric Synthetic Aperture Radar (InSAR) is an example of a successful monitoring method that has provided the opportunity of conducting local and regional surveys to detect and measure volcanic deformation. The signals detected by InSAR on volcanoes can be related to changes in pressurization of the volcanic systems, occasionally leading to volcanic eruptions.  InSAR has also been used to detect precursory inflation months prior to volcanic eruptions, proving its potential for early warning systems in some volcanoes.

There are 19 historically active volcanoes in the Philippines.  From these volcanoes, Mayon, Bulusan, and Kanlaon have erupted over the last decade.  Eruptions at these volcanoes are a threat to human life and infrastructure, having over 350,000 people living within 10 km from just these 3 volcanic centers.  For this reason, monitoring volcanic unrest in this area is of extreme importance.  

We used the ALOS-1 satellite from the Japanese Aerospace Exploration Agency (JAXA), which acquired a global L-band data set of nearly 25 acquisitions during 2007-2011, to make an InSAR time-series analysis using the Small Baseline Subset (SBAS) techinique over Mayon, Bulusan, and Kanlaon volcanoes.  We further extended the analysis over Taal caldera, due to its close proximity to the capital of Manila and because volcanic unrest has been previously reported during the time period of our analysis.  We addressed if time-dependent volcanic deformation could be detected with this data on these volcanoes.  We will present kinematic models for volcanoes exhibiting deformation to determine magma chamber depths, geometry and volume change.

Preliminary results show deformation signals on Taal caldera, Mayon, and Bulusan volcanoes, while Kanlaon volcano shows possible signs of volcanic deformation affected by atmospheric contributions.  Deformation on the volcanic complex within Taal Lake is characterized by a maximum uplift of 8 cm from 02/28/2010 to 01/16/2011, while a maximum uplift of 6 cm is observed within the same time range on the western border of Taal Lake.  Mayon volcano shows a strong subsidence signal on the southeastern flank, related to compaction of lava flows emplaced during the 2006 eruption at a rate of ~4 cm/yr.  Edifice deformation is also observed in Mayon volcano, with a maximum inflation of ~9 cm on 10/15/2007 with respect to 02/27/2007 (months prior to the 2008 eruption).  Bulusan volcano shows co-eruptive inflation of ~8 cm from 02/27/2007 to 10/15/2007 on the west-southwest part of the edifice, followed by post-eruption subsidence.  The signal at Kanlaon volcano is affected by atmospheric contributions, but when given a reference point with apparent similar atmospheric contribution the results show an average inflation rate of ~1.5 cm/yr in the western part of the edifice.    

Our InSAR time series survey reveals previously undetected volcanic deformation signals.  We observed pre-eruptive, co-eruptive, and post-eruptive deformation signals, further proving the utility of the InSAR time series technique not only as an early warning system but also for monitoring the long-term changes in volcanic behaviour.  We will further extend our analysis over all volcanoes in the Philippines in the near future.

Landslides are one of the most important geohazards affecting water reservoirs. Their activity might decrease reservoir’s volumetric capacity, damage roads, buildings and transmission infrastructure. Moreover, landslides could be harmful for the dams. Historical records, as well as numerous previous studies, show that unexpectedly triggered and rapidly moving landslides, and debris flow may initiate devastating local tidal waves.  Thus, most of the water reservoirs located in mountainous regions have detail maps and inventory of landslides that may affect coastal zones and the most hazardous areas are operationally monitored.

Polish Geological Institute - National Research Institute is responsible for the landslide mapping and monitoring of hazardous areas in Polish Carpathians. Monitoring of all slopes surrounding reservoirs is technically challenging and expensive by geodetic surveying, GNSS and borehole instrumentation (inclinometers, piezometers etc.). Therefore, remote sensing techniques and specifically the SAR interferometry, are an obvious alternative to monitor entire water reservoir and surroundings. However, mountainous areas are often very challenging for interferometric processing; mainly due to the very rough topography, sparse urbanization, dense vegetation, forest cover and relatively long winter season with the snow cover.

In order to give a give a quantitative and qualitative assessment of the applicability of InSAR for this specific application, a number of validation campaigns on pre-selected test sites have been performed. Considering the mentioned challenges in context of InSAR applicability, two landslides were selected as the optimal candidate sites. One in Poland (around Roznow Reservoir in Nowy Sacz area), and another one in the area of reservoir in the Three Gorges in China. Both reservoirs are located in mountain areas with slopes affected by landslides. Furthermore, both sites are equipped with in-situ monitoring systems, and these data are used for both validation and fusion of/with InSAR results.

In the performed study, a two-step approach was used. In the initial step, SAR Interferometry (DInSAR) technique is used to obtain a preliminary results on the areas with the high displacement rates. This initial step serves as a “localization tool”, where areas of strong deformation are localized both in terms of its location and magnitude of the signal. Because of the short revisit time TerraSAR-X data is used in this initial step. In the the second step initial DInSAR results are utilized as a-priori information and extended with additional X-band (TerraSAR-X) data analysis, and analysis of the archives of C-band (ERS 1/2 and Envisat) and L-band (ALOS) SAR data. In the second analysis step, multitemporal interferometric techniques for time series analysis to estimate the velocity of the landslides is used.

The key research questions of this study is to explore the applicability of archived SAR data (ERS-1/2, Envisat, ALOS PALSAR) and their synergy with TerraSAR-X and in-situ terrestrial measurements to improve and validate detectability of deformation signal in InSAR challenging mountainous regions. Thus, the results from InSAR analyses for both areas were examined against landslide inventory maps and the obtained deformation velocities are compared with in-situ monitoring data. It was found that the landslide activity reported by InSAR match very well with mapping inventory and it allows to fine-tune areas of contemporary activity. The comparison between in-situ monitoring data and InSAR results shows overall good match, however it need to be further explored and interpreted.

Despite this being an ongoing research, initial results will be presented in detail, results of the validation and fusion with terrestrial measurements will be extensively elaborated on, and the preliminary conclusions will be discussed.

Acknowledgments: This research is supported by ESA Dragon 3 Project, Landslide identification, movement monitoring and risk assessment over rugged mountain area using advanced earth observation techniques, project ID 10606.

Long-term time series of ground displacement are key information for the interpretation of several natural and anthropic hazards. In high mountain environments, the analysis of surface displacements and their relationship with meteo-climatic variables can be exploited to better understand the effects of climate changes on the frequency and the intensity of the landslides and their triggering phenomena (Huggel et al., 2012). In this scenario, the use of straightforward tools and approaches to analyse surface displacements at high spatial and temporal resolutions is a real need. Here we present the preliminary results of the HAMMER project, developed in the framework of NextData initiative (http://www.nextdataproject.it/). HAMMER aims at collecting ground deformation time series, and analysing eventual changes of the deformation trends associated to meteo-climatic variables. In particular, we exploited Differential SAR Interferometry (DInSAR) technique to analyse space-born measurements of surface displacements over areas characterized by slope instability and/or gravitational spreading. We used the Parallel-SBAS, namely P-SBAS (Casu et al., 2014), service recently released within the ESA’s Grid Processing On Demand (G-POD, http://gpod.eo.esa.int/) environment for the generation of Earth’s surface deformation time series and interferometric production. This service performs the full SBAS-DInSAR (Berardino et al., 2002) chain starting from Level 0 data, and generates displacement time series. Data are retrieved through the Virtual Archive 4 (http://eo-virtual-archive4.esa.int/, in the framework of Supersite initiative). The P-SBAS service within G-POD is accessible through a user-friendly web-interface, which allows selecting rapidly the input SAR data, and set easily some useful input parameters for processing (e.g., selection of temporal range, baseline and coherence thresholds set, reference point selection). We produced mean deformation, velocity maps, as well as deformation time series, on a regional scale case (Aosta Valley Region, northern Italy), and at local landslide scale (Puy landslide, Piedmont, northen Italy). The possibility to gather the final results in less than 24h (by processing an average of about 30 SAR images for each frame considered), allowed to perform in relatively short time a large number of attempts. By “tuning” the processing, we have maximized for both datasets the final coverage of coherent points where displacement time series can be traced over time. This was performed by analysing the effect of SAR images acquired in the winter season, as well as of the impact of perpendicular and temporal baseline constraints. We show here the results obtained with P-SBAS G-POD service on Valle d’Aosta region, which have been compared to the Deep Seated Gravitational Slope Deformation (DGSD) (reference IFFI project), finding a good correlation with the anomalous areas of surface deformation and the catalogued DGSD (Fig.1). In addition, the results obtained on Valle d’Aosta and Piedmont regions show a good agreement to the mean velocity maps available retrieved from the “Portale Cartografico Nazionale” http://www.pcn.minambiente.it/GN/, which was instead processed by considering PSInSAR technique on the same Envisat ASAR dataset (Fig.2). Finally, we discuss possible future developments of the P-SBAS G-POD service in the Sentinel-1 scenario, when a large amount of SAR images will be available to a greater audience, and how this may impact on the analysis of surface deformation at different spatial and temporal scales.

 References

Bernardino P., Fornaro G., Lanari R., Sansosti E. (2002) A new algorithm for surface deformation monitoring based on small baseline differential SAR interferograms. IEEE Trans. Geosci. Remote Sens., vol. 40, no. 11, pp. 2375-2382.

Casu F., Elefante S., Imperatore P., Zinno I., Manunta M., De Luca C.  Lanari R. (2014) SBAS-DInSAR Parallel Processing for Deformation Time-Series Computation. IEEE JSTARS, doi: 10.1109/JSTARS.2014.2322671.

Huggel C., Clague J. J., Korup O. (2012) Is climate change responsible for changing landslide activity in high mountains?. Earth Surf. Process. Landforms 37: 77–91.

THEME: InSAR Applications – Terrain-subsidence and landslides

KEY WORDS: Persistent Scatterers Interferometry, Non-linear Deformation, Abrupt Ground Settlement, Groundwater Extraction, Thessaly Plain

Abstract

Persistent Scatterers Interferometric (PSI) techniques are routinely applied for ground deformation measurement, providing continuous spatial coverage, that is missing in other space-borne geodetic techniques, such as GNSS and optical levelling. During the course of their development, automated processing schemes, as well as wide area coverage products were presented paving the way for an operational exploitation of the technique for geohazards monitoring. Nonetheless, there are still few cases where selection of the proper processing scheme and definition of processing parameters is required, not only to optimise performance but to allow extraction of robust results. However, due to the complexity of these application sites both in terms of environmental conditions and complexity of the deformation signal the implementation of PSI techniques is not favoured.

Thessaly plain, located at the central part of Greece mainland, could comprise such a case. The over-exploitation of the underground aquifer systems resulted in severe settlement phenomena causing on an annual basis significant damages at national infrastructures (e.g. roads and railways) and private properties as well. Although the cause to the problem is already identified, due to the severity of the consequences, the lack in understanding in depth the phenomena, in terms of magnitude, spatial extend and evolution, underlines the requirement of accurate measurements in support of decision taking and prevention actions by the local authorities.

The European Space Agency’s long SAR archive is currently the only source of available data most suitable for the extraction of the historical information  of the settlement evolution in the region and the estimation of the general trends required for management planning. However, even though numerous SAR acquisitions are available covering, the random distribution of the acquisitions in time together with the non-linearity of the deformation signal places constraints during the interferometric processing. As such, in our case the problem is mainly related to the absence of sufficient temporal sampling over the periods of severe settlement due to groundwater over-pumping  which appears as an abrupt shift in the SAR deformation histories, introducing in this way issues that have to be carefully addressed within the selected processing scheme.

Interferometric time series analysis presented herein involves the entire archive of ENVISAT for both ascending and descending orbits covering a period of almost 10 years. Processing was performed using the GAMMA software package for co-registration and geocoding purposes, as well as, for the generation of the initial differential interferograms, and StaMPS software for combined PS and SBAS-like analysis. Although the area is consisted mostly of agricultural land, sufficient number of point targets well-distributed over the entire plain was obtained, permitting the recognition of spatial variations apart from temporal changes.

Our findings outline the south part of the Eastern Thessaly plain as the mostly affected area, whereas local subsidence patterns were recognized elsewhere. Estimated mean deformation rates over the area were at the order of a few millimetres per year, though displacements up to a couple of centimetres within short time intervals of several months, in specific during summer and autumn periods, were recognized. The results were confirmed by processing independent data stacks from different tracks, highlighting the unique behaviour of the area due to intense water pumping. In addition, it is shown that displacement rate maps presenting average values over specific time spans are sometimes not sufficient and that full exploitation of the capacity of the PSI techniques by detailed examination of the displacement time series could reveal valuable information to geotechnical engineers and planners.

Finally, the vertical displacement component was calculated combining ascending and descending PSI results and the final outputs were integrated into a GIS environment for hazard susceptibility analysis following a commonly applied zonation procedure.