Large scale operational soil moisture mapping from passive MW radiometry: SMOS product evaluation in Europe & USA

Abstract

Earth Observation (EO) allows deriving from a range of sensors, often globally, operational estimates of surface soil moisture (SSM) at range of spatiotemporal resolutions. Yet, an evaluation of the accuracy of those products in a variety of environmental conditions has been often limited. In this study, the accuracy of the SMOS SSM global operational product across 2 continents (USA, and Europe) and a range of land use/cover types is investigated. SMOS predictions were compared against near concurrent in-situ SSM measurements from the FLUXNET observational network. In total, 7 experimental sites were used to assess the accuracy of SMOS derived soil moisture for 2 complete years of observations (2010–2011). The accuracy of the SMOS SSM product is investigated in different seasons for the seasonal cycle as well as different continents and land use/cover types. Results showed a generally reasonable agreement between the SMOS product and the in-situ soil moisture measurements in the 0–5 cm soil moisture layer. Root Mean Square Error (RMSE) in most cases was close to 0.1 m3 m−3 (minimum 0.067 m3 m−3). With a few exceptions, Pearson’s correlation coefficient was found up to approx. 55%. Grassland, shrublands and woody savanna land cover types attained a satisfactory agreement between satellite derived and in-situ measurements but needleleaf forests had lower correlation. Better agreement was found for the grassland sites in both continents. Seasonally, summer and autumn underperformed spring and winter. Our study results provide supportive evidence of the potential value of this operational product for meso-scale studies in a range of practical applications, helping to address key challenges present nowadays linked to food and water security.