Evaluation of two blended microwave soil moisture products with in-situ measurements and reanalysis data

The regional spatio-temporal patterns of soil moisture play a key role in climate, hydrology, food security, human health, ecosystem function, and the development and persistence of extreme weather events such as droughts, floods and heatwaves. Due to sparse and uneven in-situ measurements, it remains difficult to quantify its global distribution. Therefore, a number of global soil moisture products have been produced from different satellite sensors with different spatial and temporal resolutions. The blended microwave soil moisture products, ESA's Climate Change Initiative (CCI) and NOAA's Soil Moisture Operational Product System (SMOPS) have either better spatial or temporal coverage than other products derived from single sensors. In this study, a comprehensive evaluation of the reliability of soil moisture estimations from the CCI and SMOPS (composed of Soil Moisture Active Passive (SMAP), Global Change Observation Mission-Water (GCOM-W1), Soil Moisture Ocean Salinity (SMOS), Global Precipitation Mission (GPM), and Advanced Scatterometer (ASCAT) on EUMETSAT's Metop-A and Metop-B satellites is carried out. Resulting soil moisture products are validated with available in-situ soil moisture networks across the world, including REMEDHUS, TERENO, SMOSMANIA, RSMN, BIEBRZA_S-1, FMI, HOBI and OZNET. The period from March 2017 to March 2018 has been analyzed, giving us the opportunity to assess the satellite response to different soil moisture states under different land covers in different seasons. To remove systematic differences between satellite imagery and site-specific soil moisture data, two different scaling strategies are employed, one based on linear regression correction and the Cumulative Density Function (CDF) matching. Besides ground based measurements, ERA5 reanalysis data from the European Centre for Medium-Range Weather Forecasts (ECMWF), which has same spatial resolution as CCI and SMOPS, is used for further evaluation. The result indicates a good accuracy of CCI and SMOPS. Specifically, a good reliability of SMOPS meets the requirement for application in regions where CCI is not available. Overall, the results provide an overview of the CCI and SMOPS robustness and reliability over different areas in the world, thereby highlighting advantages and shortcomings for the effective use of these data sets for operational applications such as flood forecasting and numerical weather prediction.