Open AccessA global 1‐km downscaled SMAP soil moisture product based on thermal inertia theory
Author(s) -
Fang Bin,
Lakshmi Venkat,
Cosh Michael,
Liu PangWei,
Bindlish Rajat,
Jackson Thomas J.
Publication year - 2022
Publication title -
vadose zone journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.036
H-Index - 81
ISSN - 1539-1663
DOI - 10.1002/vzj2.20182
Subject(s) - downscaling , environmental science , radiometer , moderate resolution imaging spectroradiometer , remote sensing , normalized difference vegetation index , data assimilation , latitude , thermal inertia , water content , image resolution , meteorology , atmospheric sciences , thermal , climate change , satellite , geography , geology , precipitation , computer science , oceanography , geotechnical engineering , geodesy , aerospace engineering , engineering , artificial intelligence
Microwave remote sensing technology has been applied to produce soil moisture (SM) retrievals on a global scale for various studies and applications. However, due to the limitations of current technology, the native spatial resolution of currently available passive microwave SM products is on the order of tens of kilometers, and this resolution cannot be used to characterize SM variability on a regional scale. To overcome this limitation, a downscaling algorithm based on the thermal inertia theory–derived relationship between SM and temperature difference was developed using outputs from the Global Land Data Assimilation System–Noah Land Surface Model and the land long‐term data record–Advanced Very High Resolution Radiometer normalized difference vegetation index (NDVI) dataset and applied to the Aqua Moderate Resolution Imaging Spectroradiometer land surface temperature/NDVI data to produce a downscaled 1‐km Soil Moisture Active Passive (SMAP) radiometer daily SM product, respectively, at 6:00 a.m. and 6:00 p.m. on a global scale from 2015 to 2020. The evaluation results reveal that the downscaling model performs better in the middle or low latitudes than in high latitudes. It also performs better in warm months than in cold months. The in situ SM observations from dense networks around the world were used to validate the 1‐km and enhanced 9‐km SMAP SM data. The validation metrics indicated that both the 1‐km and 9‐km SM data have overall overestimation trends, and the unbiased RMSE (0.063 m 3 m –3 on average), mean absolute error (0.052 m 3 m –3 on average), and spatial standard deviation (0.025 m 3 m –3 on average) of the 1 km data are generally more accurate than the metrics of the 9‐km SM data, which indicates that the downscaled data provide reliable observed SM information.