Customer: Canadian Space Agency (CSA)

Programme: Radarsat

Supply Chain: CSA > CS CANADA > CS Group SPACE

Context

CS Group responsabilities for MUltisensor value-added products and Services from satellite temporal sEries analysiS are as follows:

• Datacube development.
• Sentinel-1/2 Land cover map prototype (from IOTA² library and CRIM methods).
• Flood risk assessment product.

The features are as follows:

• Optimization (and automation when possible) of Data Searching and Harvesting
• Reduction of processing times (data access, calculation,production bottlenecks mitigation…)
• Ingestion and preparation of input data to L2A level (Sen2Cor or MAJA algorithms) then ARD standard
• Data pre-processing and preparation using common reference data (Sentinel 2 by default)
• Steering functions at Datacube level: starting production, informing operators and users of data availability, management of parallelization and scalability, of processing chains, visualization of intermediate results and confidence indices, dissemination…

Project implementation

The project objectives are as follows:

• Management of the full cycle of EO data from search, collection, storage and access to dissemination (Copernicus, Radarsat/RCM…)
• Optimized processing and analysis chains and services managing volume and heterogeneity of data
• Methods for analysis and classification of radar/optical time series based on deep learning technologies and Dempster Shafer approach

The processes for carrying out the project are:

• Agile

Technical characteristics

The solution key points are as follows:

• Optimized system architecture, based on Open Source components
• Datacube: optimisation (and automation) of acquisition, pre-processing and data preparation operations => ARD data
• Machine learning for Land Use Classification of Sentinel-1/2 Data (adapted from IOTA² library)
• Dempster-Shafer approach for fusion of IOTA² and Radarsat time series land cover maps

The main technologies used in this project are:

Domain	Technology(ies)
Hardware environment(s)	High Performance Computing (cluster ready)
Operating System(s)	Unix, Linux_Ubuntu, CentOS
Programming language(s)	Python, C++
Interoperability (protocols, format, APIs)	OpenSearch Geo & Time, WMS, WPS, WCS, CLI, SDK Python, REST Interfaces
Main COTS library(ies)	Jupyter notebook, Open Datacube (ODC), OTB, GDAL, PostgreSQL, Xarray, Dask