Downscaling SMOS brightness temperatures to produce higher-resolution soil moisture analyses

With the increasing trend toward higher resolution land-surface modeling, the need to generate higher-resolution soil moisture products is becoming more evident. This study will examine the potential to produce high-resolution (5-10 km) soil moisture analyses from lower-resolution passive radiometer L-band SMOS brightness temperatures. Utilizing the Canadian Land Data Assimilation System (CaLDAS) developed at Environment Canada, the downscaling approach will focus upon high-resolution land-surface modelling to provide the first-guess soil moisture. The land-surface model will be forced with accurate, high-resolution estimates of the land-surface properties including soil texture and vegetation properties.

Within CaLDAS the lower-resolution SMOS (Soil Moisture Ocean Salinity) brightness temperature information will be downscaled within the Ensemble Kalman Filter (EnKF) methodology to update the higher-resolution first-guess soil moisture fields. The intensively observed SMAPVEX12 field campaign of 2012 provides a unique ground truth dataset to verify the higher-resolution soil moisture analyses produced. This approach differs and can be compared with alternative deterministic approaches based on the usage of higher-resolution auxiliary data (e.g., surface temperature and NDVI) from satellites such as MODIS for example.