Appraisal of SMOS soil moisture at a catchment scale in a temperate maritime climate

Soil moisture is one of the important variables in hydrological modelling, which is now possible to be measured with remote sensing. This study is an attempt to evaluate the Soil Moisture and Ocean Salinity (SMOS) satellite derived soil moisture for hydrological applications at a catchment scale. The Soil Moisture Deficit (SMD) derived from a Probability Distribution Model is used as a benchmark for all comparisons. Three approaches are used for the evaluation of SMOS soil moisture. The first approach is based on ROSETTA pedotransfer functions (PTFs), while second and the third are based on linear/non-linear and seasonal algorithms particularly for growing and non-growing seasons respectively. The field capacity and permanent wilting point estimated from the simulated Water Retention Curve (WRC) through ROSETTA are used for the transformation of SMOS data into SMD. The growing seasons used in this study belong to the months from March to November, while the non-growing seasons comprise of months from December to February. The highest performance is given by a combined growing and non-growing season algorithms with the Nash Sutcliffe Efficiencies (NSEs) of 0.75 and RMSE = 0.01 m(3)/m(3) followed by the linear and non-linear algorithms (NSE = 0.40-0.42; RMSE = 0.02 m(3)/m(3)). The worst performance is revealed by the PTFs indicating that it should be used with caution for direct coarse scale SMOS applications (NSE = -24.98 to -40.23) and need more treatments regarding the spatial and depth wise mismatch. The overall analysis reveals that SMOS soil moisture is of reasonable quality in estimating Soil Moisture Deficit at a catchment level with a local adjustment algorithm combining growing and non-growing seasons. (C) 2013 Elsevier B.V. All rights reserved.