Improving soil moisture prediction of a high-resolution land surface model by parameterising pedotransfer functions through assimilation of SMAP satellite data

Abstract. Pedotransfer functions are used to relate gridded databases of soil texture information the hydraulic and thermal parameters land surface models. The within these pedotransfer uncertain calibrated through analyses point samples. How calibrations at spatial scale modern models is unclear because represent an area average. We present a novel approach for calibrating such improve model moisture prediction by using observations from Soil Moisture Active Passive (SMAP) satellite mission data assimilation framework. Unlike traditional calibration procedures, always takes into account relative uncertainties given both observed estimates find maximum likelihood estimate. After performing procedure, we improved heat flux Joint UK Land Environment Simulator (JULES) (run 1 km resolution) when compared cosmic-ray monitoring network (COSMOS-UK) three tower sites. resolution COSMOS probes much more representative grid than point-based sensors. For 11 neutron located across modelled domain, average 22 % reduction in root mean squared error, 16 unbiased error increase correlation after techniques retrieve new function parameters.