SUPSAR-Snow
Preparing Europe’s next SAR missions to better measure snow
Seasonal snow is changing fast. SUPSAR-Snow develops and validates multi-frequency SAR methods, enabling future missions like ROSE-L and Sentinel-1 FG/NG to deliver reliable snow information, including changes in snow water equivalent (ΔSWE), SWE itself, snow depth (SD), and melt/refreeze phases. We fuse C and L band (with support from X/Ku) and package the results into open software and datasets that missions and services can use right away.
What we deliver
- ΔSWE from multi-channel InSAR (C+L): combining interferometric measurements to reduce the impact of phase wraps and decorrelation.
- SWE from backscatter (X/Ku): SMRT-based forward modeling constrained by SVS-2/Crocus and multi-frequency decoupling of soil and snow signals.
- Snow depth from DEM differencing: precise mutual calibration of bistatic/repeat-pass InSAR DEMs and deep-learning compensation for radar penetration bias (TanDEM-X → transfer to C/L).
- Melt / refreeze mapping: C-band wet-snow detection plus multi-frequency refreezing-depth retrieval refined using tower radar data.
- Open outputs: a Python Algorithm Software Package, an Input Reference Dataset Collection, experimental products, validation reports, and a Scientific Roadmap with acquisition/phase recommendations.
Why it matters
- Water & hazards: sharper streamflow and flood outlooks from improved snow mass estimates.
- Weather & climate: better initial snow states for models and assimilation.
- Mission readiness: practical guidance for Sentinel & ROSE-L tasking and higher-level product design.
How it works (in brief)
- Consolidate science requirements & sites across tundra, alpine, boreal and agricultural snow.
- Build a shared dataset from towers (SodSAR, WBSCAT), airborne (CryoSAR, UAVSAR, F-SAR), and spaceborne (Sentinel-1, ALOS-2/4, TanDEM-X, RCM…).
- Develop & validate algorithms for ΔSWE (InSAR), SWE (backscatter), SD (DEM differencing), and melt/refreeze depth; quantify uncertainties.
- Deliver software & products and compare against state-of-the-art.
- Publish, demonstrate, recommend: roadmap for future Sentinel/ROSE-L snow products and acquisitions.
Where we work
Antarctic Peninsula & ice shelves — homogeneous areas for refreezing depth tests.
Sodankylä, Finland — dense in-situ snow & tower radars.
Austrian Alps — complex terrain testbed.
Trail Valley Creek (NWT, Canada), Powassan (ON, Canada) — varied snow/soil regimes.