Assessment of the NASA MERRA-2 Climate Reanalysis and ESA CCI Satellite Remote Sensing Soil Moisture Products Over the Contiguous United States

Abstract:

This study assesses the daily NASA Modern-Era Retrospective analysis for Research and Applications, version 2 (MERRA-2) climate reanalysis and ESA Climate Change Initiative (CCI) remote sensing soil moisture (SM) products against Soil Climate Analysis Network (SCAN) in-situ measurements in nine climate regions for three year (i.e., 2014-2016). These climate regions cover the entire Contiguous United States (CONUS) and include North West (NW), West North Central (WNC), East North Central (ENC), Central (C), North East (NE), West (W), South West (SW), South (S), and South East (SE). The NASA MERRA-2 and ESA CCI soil moisture products show different accuracies in various climate regions and seasons. Over the 3-year study period, the MERRA-2 and CCA SM products have the highest and lowest accuracy in the NE and ENC regions, respectively. The highest (lowest) agreement between the MERRA-2/CCI SM estimates and SCAN SM in-situ measurements are observed in April (February) with the unbiased root-mean-square-difference (ubRMSD) of 0.0638 (0.0914). In all of the climate regions, the CCI SM data are more accurate than those of MERRA-2 in April-October. The results also show that the CCI SM data are superior than those of MERRA-2 in the NW, WNC, C, S, and W climate regions. These outcomes enhance our understanding of the performance of MERRA-2 and CCI SM products in various climate regions and seasons. The superiority of each of these products in different seasons and climate regions motivated us to generate a new SM product. The new product is created by the synergistic combination of the more accurate SM data (from MERRA-2 or CCA) in each region and season. The generated SM dataset (ubRMSD=0.0614) outperforms the MERRA-2 (ubRMSD=0.1013) and CCI SM (ubRMSD=0.0977) products.