Using Ranked Probability Skill Score (RPSS) as Nonlocal Root-Mean-Square Errors (RMSEs) for Mitigating Wet Bias of Soil Moisture Ocean Salinity (SMOS) Soil Moisture

Abstract

To mitigate instantaneously evolving biases in satellite retrievals, a stochastic approach is applied over West Africa. This stochastic approach independently self-corrects Soil Moisture Ocean Salinity (<small>SMOS</small>) wet biases, unlike the cumulative density function (<small>CDF</small>) matching that rescales satellite retrievals with respect to several years of reference data. Ranked probability skill score (<small>RPSS</small>) is used as nonlocal root-mean-square errors (<small>RMSEs</small>) to assess stochastic retrievals. Stochastic method successfully decreases <small>RMSEs</small> from 0.146 m3/m3 to 0.056 m3/m3 in the Republic of Benin and from 0.080 m3/m3 to 0.038 m3/m3 in Niger, while the <small>CDF</small> matching method exacerbates the original <small>SMOS</small> biases up to 0.141 m3/m3 in Niger, and 0.120 m3/m3 in Benin. Unlike the <small>CDF</small> matching or European Centre for Medium-Range Weather Forecasts (<small>ECMWF</small>) Re-Analysis (<small>ERA</small>))–interim soil moisture, only a stochastic retrieval responds to Tropical Rainfall Measuring Mission rainfall. Based on the effects of bias correction, RPSS is suggested as a nonlocal verification without needing local measurements.