Improving 1km Sentinel-1 Soil Moisture Retrievals by Optimizing Backscatter Preprocessing Workflows

• DOI: 10.5194/egusphere-egu23-7441
• Publication Date: 2023-02-25T20:12:41Z
• Authors: Wolfgang Wagner; Samuel Massart; Bernhard Raml; Raphael Quast; Pavan Muguda Sanj...; Claudio Navacchi; Felix Reuß; Bernhard Bauer-Ma...; Mariette Vreugdenhil
• Keywords: Backscatter (email)

Most scientific studies dealing with the retrieval of soil moisture data from Synthetic Aperture Radar (SAR) focus on formulation, training, and validation models used to convert backscatter measurements into data, while paying little attention how are preprocessed. This is insofar surprising given that topography Earth surface in combination variable SAR imaging geometry may introduce strong orbit-related geometric effects obscure signal time series. Furthermore, mechanisms characterized by a very high spatial variability, leading sensitivity moisture. Differences hardly ever accounted for except masking some obvious soil-moisture-insensitive areas such as water bodies, dense forest urban areas. In this contribution we give an overview ongoing efforts at TU Wien develop Sentinel-1 preprocessing workflows produce 1 km series optimized task retrieving same resolution. The following topics addressed: (i) use radiometric terrain corrected instead standard ground range detected products, (ii) subsurface scattering areas, other (iii) standardization reference incidence angle using machine learning techniques. Our preliminary results over Europe Mediterranean region show substantial improvement retrievals would be impossible achieve sole algorithm.AcknowledgementsWe acknowledge funding European Space Agency (DTE Hydrology 4DMED), Copernicus Land Monitoring Service, Austrian Applications Programme (ROSSHINI GHG-KIT). computational presented have been achieved part Vienna Scientific Cluster (VSC).