A global comparison of alternate AMSR2 soil moisture products: Why do they differ?

This study assesses two remotely sensed soil moisture products from the Advanced Microwave Scanning Radiometer 2 (AMSR2), a sensor onboard the Global Change Observation Mission 1 - Water (GCOM-W1) that was launched in May 2012. The soil moisture products were retrieved by the Japan Aerospace Exploration Agency (JAXA) algorithm and the Land Parameter Retrieval Model (LPRM) developed by the VU University Amsterdam, in collaboration with the National Aeronautics and Space Administration (NASA). The two products are compared at the global scale. In addition, the products are evaluated against field measurements from 47 stations from the COsmic-ray Soil Moisture Observing System (COSMOS) network which are located in the United States (36 stations), Australia (7 stations), Europe (2 stations) and Africa (2 stations). After examining the retrieval algorithms, it is hypothesized that four factors, namely, physical surface temperatures, surface roughness, vegetation and ground soil wetness conditions, may affect the quality of soil moisture retrievals. From the inter-comparisons at the global scale, the correlations of the two products highlight differences in the representation of the,seasonal cycle of soil moisture, with negative correlations found for several regions. Correlations of the anomaly time series were generally strong (R > 0.6) as a result of soil moisture sensitivity to external meteorological forcing and possibly also random noise in the satellite observations. Due to the inherent differences in spatial coverage and measurement scale of the COSMOS and satellite data, the comparisons in terms of correlation coefficients are the most reliable. It was found that both products show rapid decreases in correlation coefficients under low mean temperature (<290 K), high mean EVI (>0.3) and highly wetted conditions. These findings are further supported by the bias and RMSE estimates which show that JAXA has relatively better performance under dry conditions while the bias and RMSE of LPRM are generally smaller than JAXA, when considered against the four variables. These results provide information on appropriate parameterizations and model selection for the retrieval algorithms and a future research direction to improve the quality by leveraging the strengths of-the JAXA and LPRM algorithms. With these, when a multi-year dataset is available, there will be more confidence in defining the seasonal cycle and the data can be decomposed to identify the anomalies where the bias is not relevant. (C) 2015 Elsevier Inc. All rights reserved.