Establishing soil moisture downscaling parameters a priori using network data as a proxy for future RADARSAT-Constellation retrievals

English Abstract:

Gridded soil moisture datasets are increasingly available over agricultural regions, including those from the Soil Moisture Ocean Salinity (SMOS) and the recently launched Soil Moisture Active Passive (SMAP) missions. These products are available at relatively coarse resolutions (e.g. > 9 km) and methods are still needed to downscale these data to field scale. C-band SAR modelling can provide distributed soil moisture maps to establish soil moisture scaling distributions within coarse footprints, yet this approach is only feasible in the period prior to vegetation growth. Furthermore, the overpass frequency of current sensors limits the number of observations from which these distributions of soil moisture can be acquired. The future RADARSAT-Constellation will overcome this limitation using 3 satellites thus enabling field scale soil moisture mapping every few days. These frequent retrievals can be used to construct a priori information on the soil moisture scaling distribution, and its time stability to constrain posterior downscaling functions during the growing season. In this study, data from the Brightwater Creek in-situ soil moisture network are used as a proxy of SAR modelled soil moisture to evaluate this proof-of-concept and its applicability over the growing season.