Session: Land I

The Copernicus Global Land service provides continuously a set of bio-geophysical variables describing, over the whole globe, the vegetation dynamic, the evergy budget at the continental surface and some components of the water cycle. These generic products serve numerous applications such as agriculture and food security monitoring, weather forecast, climate change impact studies, water, forest and natural resources management.

The portfolio of the Copernicus Global Land service contains Essential Climate Variables like the Leaf Area Index (LAI), the fraction of PAR absorbed by the vegetation (FAPAR), the surface albedo, the Land surface Temperature, the soil moisture, the burnt areas, the areas of water bodies, and additional vegetation indices. They are generated every hour, every day or every 10 days on a reliable and automatic basis from Earth observation satellite data. Beside this timely production, the available historical archives have been processed, using the same innovative algotirhms, to get consistent time series as long as possible. All products are accessible, free of charge after registration through FTP/HTTP (http://land.copernicus.eu/global/) and through the GEONETCast satellite distribution system. The products are provided with documents describing the physical methodologies, the technical properties of products, and the results of validation exercises. The service performs a continuous quality monitoring, complemented by regular reviews done by independant experts and feedback from a technical user group.

For the service continuity, research and development are performed on two parallel ways. On one hand, the existing retrieval methodologies are being adapted to PROBA-V 1km resolution data. On the other hand, the service is moving the operations to finer resolution (300m). This is prepared in two ways: (1) the FP7/ImagineS project focuses on the retrieval of LAI, FAPAR, fraction of vegetation cover and surface albedo; (2) the evolution activities in the context of the Global Land service focusing on the other variables of the portfolio. So far, both use the PROBA-V sensor data. The operational production of LAI, FAPAR and fraction of vegetation cover at 300m resolution over Europe is being to start soon. The spatial coverage will be extended to the globe by the end of the year while the production of other variables (Albedo, NDVI, Burnt Areas, Water Bodies) will begin.

In the next step, the methodologies and the processing chains will be adapted to ingest the Sentinel-3 land data, jointly with the PROBA-V data. Thus, they will provide a better spatial and temporal coverage of the continental ecosystems. However, for a near real time production, it is mandatory that the Sentinel-3 Level 2 synergy land products are delivered in less than 48 hours after the acquisition so that the final Global Land products can be disseminated to users according to the service specifications i.e. within 3 days after the imagery acquisition.

The talk will present the status and the evolution of the Global Land Service, and will show how Sentinel-3 can contribute to a future 300m land monitoring service.

At the end of 2015, Sentinel-3 will be launched. With its two instruments OLCI (Ocean Land Colour Instrument) and SLSTR (Sea and Land Surface Temperature Radiometer), the successor of the Envisat MERIS sensor will allow ensuring the continuity of global land cover maps production initiated in the CCI Land Cover (CCI-LC) project.

Started in 2011 in the framework of the ESA Climate Change Initiative (CCI), this project aims at delivering global land cover products that match the needs of key users from the climate modelling community. Land cover is one of the Essential Climate Variables (ECVs) listed by the Global Climate Observing System (GCOS). It refers to one of the most obvious and commonly used indicators for land surface parameterization and is much associated to the human activities and natural processes, while also playing a significant role in climate forcing.

At the beginning of the project, an analysis of the specific needs of this community revealed the priority for stable and consistent land cover products over time. It was indeed not the case with existing products, such as the GlobCover or MODIS ones, where significant year-to-year variations are observed which are not due to land cover changes but rather to classifications instabilities.

To meet this requirement, the land cover concept was revisited to decouple the stable and dynamic component of the land cover. It was proposed to map the stable component of the land cover using a multi-year classification strategy and then, to provide separate additional information about the land surface natural variability. It was indeed found that classification instabilities are located in areas known to be either heterogeneous and/or showing contrasted seasonal cycles. Using a multiyear Earth Observation dataset can contributed to reducing the classification sensitivity to the period of observation and then, to better extracting the stable land cover component.

In October 2014, the project delivered a set of three consistent global land cover products corresponding to the 1998-2002, 2003-2007 and 2008-2012 periods, along with climatological 7-day time series representing seasonal dynamics of the land surface (in terms of vegetation greenness, burned area and snow).  

The 3 maps were obtained by processing, as main input, the full archive (2003-2012) of MERIS full resolution (300m) instrument. MERIS reduced resolution and SPOT-Vegetation time series were also used as complementary dataset. This huge amount of data was processed to generate a single map over the 2003-2012 period. This unique map was used as a baseline to derive the 3 output products: it was back- and up-dated at the location and time major land cover changes were detected.  The accuracy of the 2008-2012 global LC map was assessed using the GlobCover 2009 validation dataset, resulting in a weighted-area overall accuracy of 74.1% which is slightly better than previous products. In addition, visual analysis revealed a much better delineation of landscape patterns.

Since the Envisat mission suddenly stopped in 2012, the intention is to make the coming Sentinel-3 the new key sensor of the whole CCI-LC system. In the medium term, Sentinel-3 is indeed the only sensor that can ensure continuity in the global acquisition of medium spatial resolution time series on daily intervals. Launched in May 2013, PROBA-V, which is the SPOT-Vegetation successor, is currently making the job and will also be used at its maximum potential in the project. Yet, even if the mission is up-to-now successful, its official operational lifetime is set to only 2.5 years.

Concretely, waiting for Sentinel-3, the project will rely on PROBA-V time series. Consistency with previous product will be possible through an optimal exploitation of the 1-year overlap between SPOT-Vegetation and PROBA-V. Then, the project will integrate Sentinel-3 time series in higher and higher quantity as soon as they are available. Looking for synergies between the two sensors will be a key target.

An important contribution for the consistency of the Sentinel-3 time series across the other sensors is a consistency of the surface reflectance values as well as best possible determination of the magnitude of their uncertainties calculated through the sensor-specific pre-processing chain. Conversely, several aspects make this consistency objective more challenging. First, the fact that different band settings used for each sensor considered here has an effect on the used pre-processing modules on one hand and on the accuracy of the surface reflectance values on the other hand. Second, some information critical for the pre-processing cannot be retrieved with an acceptable accuracy - or worse, cannot be calculated from the image information itself - and therefore comes from external sources. Furthermore, the harmonization of all auxiliary data (like DEM, land water mask, atmospheric data) has a significant impact on the consistency of the time series. Finally, the geometric accuracy of the different sensors has to be investigated with the aim of knowing more about the inter-sensor compatibility.

Addressing all these challenges, the CCI-LC project should be able to deliver in 2016-2017 new global LC maps at 300m spatial resolution to cover the period after 2012 and provide the climate modelling community with around 15 years of consistent land cover mapping.

The World Food Program is the largest humanitarian agency in the World assisting close to 90 million people across 75 countries. The majority of WFP's of beneficiaries are rural poor farmers and pastoralists in marginal areas and climate related factors are the key driver of their humanitarian needs.

In order to support its operations and to better design its interventions, WFP is making increasing usage of Earth Observation data in particular medium resolution data such as vegetation indices, rainfall estimates, snow cover and LST. These are used both in an early warning context and as input to country or regional level climate-environment analysis.

Currently WFP's Seasonal Monitor provides coverage of growing season conditions for all the agency's areas of operation, which implies global coverage minus Europe, North America, Australia and polar regions. MODIS data provides the primary input for this monitoring system, yielding vegetation index and snow cover anomalies, coupled with rainfall estimates from the Climate Hazards Group at the University of California - Santa Barbara. The system's main outputs are advance warning of drought conditions which are then integrated with a wealth of household level information in order to provide assessments of likely impacts on food security of poor vulnerable populations.

Climate-environmental analysis identifies and maps patterns of interannual variability, changes in seasonality (onset of season, length of growing period) and medium term trends in vegetation cover, rainfall and snow cover as well as evaluating frequency of hazards such as drought and floods. These are nowadays essential components in WFP planning helping to define target areas for interventions as well as the nature and timing of the intervention. Most of these analysis are based on medium and long term records of satellite data such as MODIS, GIMMS AVHRR NDVI, SPOT-VGT.

Sentinel-3 data offers both new possibilities and a garantee of continuity of data streams that are fast becoming essential for WFP. Current systems will be detailed and the WFP user perspective on how best Sentinel 3 could serve its operations and how best it could fit in current systems will be outlined.

Gross primary production (GPP) of terrestrial vegetation is a key component in the carbon cycle and as such an important component of the carbon exchange between the atmosphere and the biosphere. At present the only global operational GPP products is derived from MODIS data (i.e. MOD17) and with the MODIS mission reaching its end of life, there is no plan from continuity of such an important  product from other satellite mission. However, the recent development of a new diagnostic model called the Southampton Carbon flux (SCARF) model provides the opportunity to estimate GPP from MERIS data (and Sentinel 3 OLCI in future). The model, based on the quantum yield of vegetation, overcomes some of the limitations in the current MODIS product and provides a better estimate of carbon uptake by vegetation.  The model exploits the existence of distinct quantum yields in the two key photosynthetic pathways (i.e. C3 and C4) to estimate the conversion of absorbed photosynthetic radiation into GPP. Furthermore, the model uses the MERIS Terrestrial Chlorophyll Index (MTCI) as a surrogate to the fraction of photosynthetically active radiation absorbed by photosynthetic elements of the canopy (i.e. FAPAR_ps) rather than the total canopy FAPAR. The model predicated GPP was strongly correlated with in-situ GPP measured at a number of flux tower sites across Europe, North America and Africa covering major biomes. In addition the spatial distributions of the GPP across these continents were comparable to those from MOD17 GPP product. Sentinel 3 OLCI will provide continuation of the MTCI, one of the key inputs to the model. With further calibration and understanding the feasibility of operational production, the SCARF model provides the potential to develop a new global GPP product from Sentinel 3 OLCI data. This, in turn, will provide a complementary (or comparable) product to the existing MODIS products and will ensure future continuity of one of the important component to understand the functioning of ecosystem processes and their response/feedback to climate