Comparing surface-soil moisture from the SMOS mission and the ORCHIDEE land-surface model over the Iberian Peninsula

• Published in: Remote sensing of environment 2016-03, Vol.174, p.69-81
• Main Authors: Polcher, Jan, Piles, Maria, Gelati, Emiliano, Barella-Ortiz, Anaïs, Tello, Marivi
• Format: Article
• Language: English
• Subjects: Enginyeria de la telecomunicació; Humitat; Land-surface modelling; Measurement; Mesurament; ORCHIDEE; Radiocomunicació i exploració electromagnètica; Remote sensing; SMOS; Soil moisture; Sòls; Teledetecció; Àrees temàtiques de la UPC
• ISSN: 0034-4257; 1879-0704
• DOI: 10.1016/j.rse.2015.12.004

The aim of this study is to compare the surface soil moisture (SSM) retrieved from ESA's Soil Moisture and Ocean Salinity mission (SMOS) with the output of the ORCHIDEE (ORganising Carbon and Hydrology In Dynamic EcosystEm) land surface model forced with two distinct atmospheric data sets for the period 2010 to 2012. The comparison methodology is first established over the REMEDHUS (Red de Estaciones de MEDición de la Humedad def Suelo) soil moisture measurement network, a 30 by 40km catchment located in the central part of the Duero basin, then extended to the whole Iberian Peninsula (IP). The temporal correlation between the in-situ, remotely sensed and modelled SSM are satisfactory (r>0.8). The correlation between remotely sensed and modelled SSM also holds when computed over the IP. Still, by using spectral analysis techniques, important disagreements in the effective inertia of the corresponding moisture reservoir are found. This is reflected in the spatial correlation over the IP between SMOS and ORCHIDEE SSM estimates, which is poor (ρ~0.3). A single value decomposition (SVD) analysis of rainfall and SSM shows that the co-varying patterns of these variables are in reasonable agreement between both products. Moreover the first three SVD soil moisture patterns explain over 80% of the SSM variance simulated by the model while the explained fraction is only 52% of the remotely sensed values. These results suggest that the rainfall-driven soil moisture variability may not account for the poor spatial correlation between SMOS and ORCHIDEE products. •Comparison of remote sensed and simulated surface soil moisture (SSM) estimates.•Over the Iberian Peninsula the temporal correlation of SSM at point level are high.•The spatial correlation of the SMOS and simulated SSM is small throughout the year.•Singular value decomposition (SVD) is performed between rainfall and SSM estimates.•SVD shows that in remote sensed SSM a smaller part of variance is explained by rain.