Canadian Space Agency SAR-related ARD activities - PowerPoint Presentation

Slide1

Canadian Space Agency

SAR-related ARD activities

Paul Briand,

LSI VC 3 meeting

March 20-21, 2017

Paul.briand@canada.ca

Context

The CSA, in partnership with different Departments, is starting five new initiatives to ensure an increasing use of space-based EO data in key policy areas of terrestrial monitoring activities supporting climate change impacts and ecosystem resilience. Multi-sources EO dataInteroperability and complementarityDevelopment of EO derived productsLand cover and land cover changeEssential ecosystem variablesDevelopment of geospatial big data processing/analytics toolsData cube analytic environment Optical and SAR ARDToolboxes enabling full exploitation of EO dataGenerate derived products and information in support of evidence-based decision making for ecosystems monitoring and climate change

2

|

SAR - ARD

NRCan-CCMEO initiative:Development of a big geospatial data analytics (BGDA)systems, referred as the Geoanalytics Earth Observation Data Environment (GEODE).Objectives:Conduct scientific planning and outreach to develop a shared cross-sectoral vision to identify BGDA needs & requirements and which develops opportunity for collaborative R&D partnerships in the academic and commercial sectors to address these needs & requirementsEstablish the requirements of ARD from the Radarsat Constellation Mission and from optical sensors Improve capacity to characterize Canada’s landmass through multi-sensor fusionDevelop capacity to Develop a toolbox to generate on-demand value-added products depicting vegetation change characteristics across Canada Specific workpackage on developing and assessing SAR and RCM based ARD for time series analytics

3

|

SAR – ARD WorkplanPhase 1

Development of a specification for RADARSAT-based ARD considering the GEODE environment and interoperability with other EO datasets. This will include examining current SAR ARD standards/definitions under development (e.g. CEOS) and operational (e.g. Google Earth Engine) and the development of mini-cubes to support development, testing and demonstration.First activity: Users needs and requirements for SAR ARDBig Geopatial Data Analytics Workshop, Ottawa, January 30, 2017Moving Towards Analysis Ready SAR Data

4

|

SAR ARD Workshop Outcomes

The characterization of initial radar-‐related ARD requirements and analytics represented a challenging task.Different SAR User groups and applications have different SAR Data and product needs.RadiometryPreference is more for a Beta-0 or Sigma-0 projection with an incident angle layer as additional channel to the product. Most numerical models require Sigma-0 and incident angle information as input. Local slope correction should not be imposed in ARD, since DEM layer quality in the cube can/will evolve over time (TerraSAR-X and Lidar DEM). Local slope correction is low processing cost, so it can be easily done on the fly during the cube analysis.Incident angle fileAn incident angle channel would be added to the product date in order to enable numerical modeling.

5

|

SAR ARD Phase 1: Workshop Outcomes

PhaseKeep and enable use of the phase information of the radar signals (InSAR applications).Polarization:Preservation of the polarimetric phase through the processing chain to ARD (i.e speckle filtering, orthorectification).CompressionNo real discussion on data compression yet (8-16-32 bits). For 32 bits, it might be a real issue depending on the cube infrastructure and available resources.

6

|

SAR – ARD WorkplanNext steps

Phase 2:Development of an ARD production roadmap to support the ingest and staging of these data in the GEODE. This will include determining scope of information requirements required to support time series analysis – (e.g. amplitude only or amplitude+phase)Phase 3:Research and development of analytical tools capable of utilizing radar-based time series and where advantageous, integrating with other data available in the GEODE.

7

|