• 14:40 SAR Tomography imaging of a tropical forest at P- and L-band

  Ho Tong Minh, Dinh (1); Le Toan, Thuy (1); Tebaldini, Stefano (2); Rocca, Fabio (2) 1: CESBIO, France; 2: DEIB, Polimi, Italy

  Show abstract

  The objective of this paper is to provide a better understanding of tomographic capabilities in characterization of dense forested areas at P-and L-band. The analysis is carried out on airborne data acquired by ONERA over the site of Paracou, French Guyana, during the ESA campaign TropiSAR. The results shown support the idea that ground- and -volume interactions play a negligible role at L-band, whereas they are significant at P-band. For a dense forest of 30 m and more, there is no ground contribution at L-band. The L-band tomographic profile is quite disturbed as compared to the P-band profile in dense tropical forest areas. In this condition, the use of tomographic imaging at L-band in tropical forests appears limited. However, when the forest top height is roughly below 20 m (e.g., in forest regrowth), the tomographic results are expected to be the same as in boreal forests. Whereas P-band tomography allow us to retrieve the whole forest vertical structure, better characterizing of the ground and/or volume scatterings and providing an unique solution in high biomasss ranging from 150-600 t/ha.

  
   

• 15:00 Airborne Holographic SAR Tomography at L- and P-band

  Ponce, Octavio; Moreira, Alberto German Aerospace Center (DLR) - Microwaves and Radar Institute, Germany

  Show abstract

  As a part of the analysis of future SAR space missions, experimental airborne realizations have been carried out to understand the 3-D backscattering of natural scenarios, such as forests. Advanced SAR imaging modes, like polarimetric Interferometric SAR (POLinSAR) and SAR tomography (SARTom) rely mostly on a single or a few angular measurements, so that the estimation of bio-geophysical parameters, such as extinction factor or biomass is retrieved by exploring only a small fraction of complete observation space. As a solution, new real data collections have been carried out in holographic SAR tomography (HoloSAR) imaging mode, which consists of the synergy of a circular and a vertical synthetic apertures. In this respect, this work presents an analysis and comparison of the results of two airborne SAR campaigns at L- and P-band with the DLR’s F-SAR sensor in quad-polarization operation. It is shown that this mode provides added benefits, such as very high resolution over 360 degrees in the azimuth-range space, while retrieving the complex reflectivity in the vertical direction. Due to this fact, uncertainties in the assessment of biosphere variables can be reduced, and the factor of angular dependency can be further understood. Also, as part of the study, pros and cons of linear and non-linear focusing methods for this mode are discussed. It encompasses the fast factorized back-projection (FFBP), and compressive sensing (CS) approaches for several constellations in elevation. Special emphasis is placed on the investigation of potential HoloSAR applications and possible spaceborne mission concepts.

  
   

• 15:20 TROPICAL FOREST BIOMASS RETRIEVAL USING P-BAND POLTOMSAR DATA

  El Hajj Chehade, Bassam (1); Ferro-Famil, Laurent (1); Ho Tong Minh, Dinh (2); Thuy, Le Toan (2); Stefano, Tebaldini (3) 1: university of rennes 1, France; 2: Centre d’Etudes Spatiales de la Biosphère (CESBIO); 3: Politecnico di Milano

  Show abstract

  Forested areas represent one-third of earth surface and are a major source of carbon storage. During the last thirty years, deforestation rate has been increasing continuously leading to an augmentation of carbon dioxide (CO2) percentage in the atmosphere. Thus, the mapping of forest biomass becomes urgently needed to better evaluate the role of forest as carbon source and sinks. [1]. PolTom- SAR techniques allow forest vertical structure characterization [2]. Thereby, several parameters may be accurately retrieved. Some of those parameters can be related to AGB [3]. This paper investigates the relation between forest parameters, retrieved with PolTomSAR techniques, and AGB. Robust estimation of forest parameters is required for biomass mapping. RVoG model, introduced by Cloude and Papathanassiou [4], is an efficient tool for forest structure characterization. A Multi-Baseline (MB) generalization of RVoG model allows the decomposition of the PolTom covariance matrix as a Sum of Kronecker Products (SKP) [5]. This decomposition leads to an accurate spatial and polarimetric separation of ground vegetation contributions [6]. Afterward, a modeling of vegetation vertical structure function is required for forest parameters retrieval. The efficiency of constant extinction vegetation model is shown in [2]. This model allows the estimation of several parameters like linear extinction, vegetation height and normalized volume intensity [2]. The aim of this paper is the study of the relation between those parameters and biomass. Moreover, a comparison with the two most common methods of biomass estimation is made. The first method consists of biomass estimation with allometric relation. While the second one, introduced by Ho Tong Minh et al. [3], is based on the strong correlation between biomass and 30 meters-above ground tomographic measured backscattered power. Besides, a physical interpretation of this method is made. The analysis of tropical forests over the test site of Paracou, French Guiana, demonstrates the potential of RVoG model on forest structure characterization at P-band. References [1] Le Toan, T. and Quegan, S. and Davidson, M. and Blazter, H. and Paillou, P. and Papathanassiou, K. and Plummer, S. and Rocca, F. and Saatchi, S. and Shugart, H. and Ulander, L., “The BIOMASS mission: Mapping global forest biomass to better understand the terrestrial carbon cycle”. Remote Sens. Environ., vol. 115, no. 11, pp.2850-2860, 2011. [2] El Hajj Chehade, B. and Ferro-Famil, L., “PBand Polarimetric SAR tomography for tropical forest structure characterization”, in EURAD 2014. [3] Ho Tong Minh, D. and Le Toan, T. and Rocca, F. and Tebaldini, S. and Mariotti d’Alessandro, M. and Villard, L. , “Relating P-Band Synthetic Aperture Radar Tomography to Tropical Forest Biomass”. IEEE Transactions on Geoscience and Remote Sensing, vol. 52, no. 2, pp.967-979, 2014. [4] Cloude, S.R. and Papathanassiou, K.P., “Three-stage inversion process for polarimetric SAR interferometry”. IEEE Proc.-Radar Sonar Navig., vol. 150, no. 3, pp.125-134, 2003. [5] Tebaldini, S., “Algebraic Synthesis of forest scenarios from Multibaseline PolInSAR Data”. IEEE Trans. on Geosc. and Remote Sens., vol. 47, pp. 4132-4142, 2009. [6] Tebaldini, S., “Multi-Baseline Polarimetric SAR Tomography of a boreal forest at P- and L-bands”. IEEE Trans. on Geosc. and Remote Sens., vol. 50, pp. 232- 246, 2012.

  
   

• 15:40 Forest Structure Information Retrieval from Polarimetric Tomographic SAR Data

  Tello, Marivi (1); Heym, Michael (2); Fischer, Rico (3); Pardini, Matteo (1); Papathanassiou, Konstantinos (1) 1: DLR, Germany; 2: Technical University Munich (TUM), Germany; 3: Helmholtz Centre for Environmental Research (UFZ), Germany

  Show abstract

  In the last years, the potential of SAR tomography (TomoSAR) techniques at low frequencies for 3D imaging of the forest has been demonstrated in the framework of several airborne experiments. However, research is still ongoing on how to fully assess the sensitivity of tomographic measurements to structural properties of the forest and on how to translate them into geophysical information of interest. In ecology, it is commonly assumed that forest structure is the vertical and horizontal arrangement of the different plants and tree elements. It is usually quantified by individual based parameters, such as stand density in the horizontal dimension and species profile in the vertical one [1]. Since any individual based parameter cannot be directly measured by the TomoSAR system, different variables have to be considered to reflect analogous information. Following this motivation, two descriptors estimated from the TomoSAR single polarization reflectivity profiles were recently proposed to account for the simultaneous spatial variability of the forest tomographic patterns, both in the horizontal and the vertical dimensions [2]. Essentially, the horizontal structure descriptor aims at measuring the similarity of the vertical reflectivity profiles in a given spatial neighborhood, whereas the vertical structure descriptor accounts for the number of layers (assimilated here to meaningful peaks) of the reflectivity profile. In this paper, these descriptors are extended to consider polarimetric TomoSAR information. First, a full rank polarimetric inversion based on Compressive Sensing is carried out. This results in an estimation of the 3D coherency matrices of the observed scene. Then, for contiguous spatial resolution cells, the similarity of the corresponding distribution of coherency matrices as a function of height can be measured, based on Wishart distances. The capability of the descriptors proposed to reflect forest structure as it is approached in ecology has been tested in the framework of the Helmholtz Alliance [3]. On the one hand, the TomoSAR based forest structure descriptors were estimated on a set of multibaseline polarimetric data acquired at L-band by DLR’s E-SAR airborne sensor, over the area of Traunstein, in Germany. The observed forest is constituted by managed forest stands, with different species composition (coniferous, deciduous or mixed forest) at different growth stages, in typical temperate conditions. On the other hand, the parameters usually employed to measure forest structure in ecology were evaluated from in situ measurements, in a set of around 100 inventory plots where the number and spatial distribution of the trees in an area of 20 x 20 meters, as well as some of their morphological characteristics, are available. The forest structure descriptors based on TomoSAR measurements and the ones based on in situ observations are compared and a good agreement is reached. References [1] M. Snyder. What is Forest Stand Structure and How Is It Measured? Northern Woodlands, Issue No. 64, pp.15, Spring 2010. [2] M. Tello, M. Pardini, K. Papathanassiou, “Towards Forest Structure Characteristics Retrieval from SAR Tomographic Profiles”, Proceedings on EUSAR, June 2015. [3] HGF Alliance website: http://hgf-eda.de/

  
   

Highlights

In May 2013, the Biomass mission concept was selected to become the next in the series of satellites developed to further our understanding of Earth.

Released: 21/06/2013

REPLAY EARTH EXPLORER 7 USER CONSULTATION MEETING (PART 1)

Released: 11/03/2013

POLinSAR 2013 Workshop

Released: 22/02/2013