Error characterization of SMOS, ASCAT, SMAP, ERA and ISMN soil moisture products: Automatic detection of cross-correlation error through extended quadruple collocation

Abstract

The triple collocation (TC) technique is being increasingly used to validate soil moisture retrievals derived from different sensors. In recent years, several extensions of this method were proposed in order to evaluate the error standard deviations of more than three systems. As the number of datasets grows the TC fundamental hypothesis, i.e. the absence of cross-correlation between the system errors, could be violated. In this paper, an Extended Quadruple Collocation (E-QC) is proposed to consider the presence of the error cross-correlation between soil moisture products, identifying automatically the couple of cross-correlated systems. The method is applied to soil moisture retrievals provided by satellite (SMOS, ASCAT, SMAP), model (ERA _Interim ) and in situ probes (ISMN). The method identified the presence of error cross-correlation between the satellite products and was able to correctly retrieve the system errors, otherwise biased if cross-correlation is not taken into account.