Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Hydrology and Remote Sensing Laboratory » Research » Publications at this Location » Publication #380935

Research Project: Integrating Remote Sensing, Measurements and Modeling for Multi-Scale Assessment of Water Availability, Use, and Quality in Agroecosystems

Location: Hydrology and Remote Sensing Laboratory

Title: Impact of vegetation water content information on soil moisture retrievals in agricultural regions: An analysis based on the SMAPVEX16-MicroWEX dataset

item JUDGE, J - University Of Florida
item LUI, P.W. - National Aeronautics And Space Administration (NASA)
item MONSIVAIS-HUERTERO, A. - University Of Florida
item BONGIOVANNI, T. - University Of Texas At Austin
item CHAKRABARTI, S. - Collaborator
item STEELE-DUNNE, S. - Delft University
item PRESTON, D. - University Of Florida
item ALLEN, S. - University Of Florida
item BERMEJO, J. - Delft University
item RUSH, P. - University Of Florida
item DEROO, R. - University Of Michigan
item COLLIANDER, A. - Jet Propulsion Laboratory
item Cosh, Michael

Submitted to: Remote Sensing of Environment
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/24/2021
Publication Date: 9/1/2021
Citation: Judge, J., Lui, P., Monsivais-Huertero, A., Bongiovanni, T., Chakrabarti, S., Steele-Dunne, S., Preston, D., Allen, S., Bermejo, J.P., Rush, P., Deroo, R., Colliander, A., Cosh, M.H. 2021. Impact of vegetation water content information on soil moisture retrievals in agricultural regions: An analysis based on the SMAPVEX16-MicroWEX dataset. Remote Sensing of Environment. 265:112623.

Interpretive Summary: Soil moisture remote sensing relies upon an accurate understanding of the amount of water stored in any plant matter above the surface. To estimate this, it is necessary to use widely available remote sensing products for global estimation. These products are commonly based on visible and near infrared channels. A study was conducted in central Iowa using a ground based microwave radiometer and accuracies were assessed for estimating soil moisture based on a variety of vegetation indices, based on both visible/near infrared and microwave radar indices. It was shown that a combination of both of these vegetation products provides the most accurate soil moisture product based on a microwave signal. This study will inform new vegetation datasets which will be used to improve microwave remote sensing programs in the future.

Technical Abstract: Retrievals of soil moisture (SM) from passive microwave remote sensing depend upon the accuracy of vegetation information, particularly in agricultural regions with dynamic conditions during the growing seasons. The SMAP SM retrieval algorithm uses vegetation water content (VWC) derivedvfrom remotely sensed, climatologically-based Normalized Difference Vegetation Index (NDVI), which doesn't respond to realtime vegetation dynamics and suffers from saturation issues. In this study, impacts of using VWC derived from two optical and three microwave indices on SM retrievals are investigated. It uses high temporal resolution data from ground and aircraft during the Soil Moisture Active Passive Validation-Microwave Water, Energy Balance Experiment in 2016 (SMAPVEX16- MicroWEX) that was conducted to improve our understanding of the sensitivities of brightness temperature and backscatter to SM during a growing season of corn and soybean. Empirical relationships were determined between in situ measurements of VWC and the indices. Of the two optical indices, the Normalized Difference Water Index (NDWI)-derived VWC resulted in lower SM retrieval root mean square difference (RMSD) of 0.036 m3/m3 with in situ SM, when compared to the differences of SM retrieved using the in situ VWC and in situ SM. The SM retrievals using NDVI-derived VWC consistently provided lower estimates of SM compared to those from in situ VWC, with mean differences increasing to 0.04 m3/m3 in the late season. Among the three radar indices, vertically polarized cross-pol ratio (CRvv)-derived VWC provided similar RMSDs in retrieved SM as the NDWI-derived VWC during the growing season. The horizontally polarized cross-pol ratio (CRhh)-derived VWC provided the worst retrievals, while the radar vegetation index (RVI)-derived VWC improved in the late season, and provided RMSDs similar to the CRvvderived retrievals at 0.028 m3/m3. Since the microwaves have better canopy penetration capability and higher reliability during cloud cover, it ensures that the observations are available to capture seasonal as well as inter-annual availability. Results presented here suggest that SMAP SM retrievals could be improved through the combined use of both near-real time NDWI and CRvv -derived vegetation information. The empirical equations developed in this study can be used for estimating VWC in SMAP SM retrieval algorithms.