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 #374288

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: Estimating corn canopy water content from normalized difference water index (NDWI): an optimized NDWI-based scheme and its feasibility for retrieving corn VWC

item CHAI, L. - Beijing Normal University
item JIANG, H. - Beijing Normal University
item Crow, Wade
item LIU, S. - Beijing Normal University
item ZHAO, S. - Beijing Normal University
item LIU, J. - Beijing Normal University
item YANG, S. - Beijing Normal University

Submitted to: IEEE Transactions on Geoscience and Remote Sensing
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/15/2020
Publication Date: 12/11/2020
Citation: Chai, L., Jiang, H., Crow, W.T., Liu, S., Zhao, S., Liu, J., Yang, S. 2020. Estimating corn canopy water content from normalized difference water index (NDWI): an optimized NDWI-based scheme and its feasibility for retrieving corn VWC. IEEE Transactions on Geoscience and Remote Sensing. 1-14.

Interpretive Summary: Vegetation water content (VWC) describes the mass of water contained per unit area in a crop canopy. Estimates of VWC are particularly useful as an early indicator of agricultural drought. The earlier that drought events can be detected – the wider the range of approaches that can be employed to mitigate their impact on the productivity and financial health of farms. This paper describes a new remote sensing algorithm for detecting corn VWC using visible and near-infrared remote sensing. Results demonstrate that our new approach is superior to existing remote sensing algorithms and represents an important step forward in our ability to operationally monitor VWC over large agricultural regions. This approach will eventually be used by operational agricultural drought monitoring systems to improve the early detection of agricultural drought in the United States corn belt.

Technical Abstract: Here, four normalized vegetation water index (NDWI) variants, i.e., NDWI(860,970), NDWI(860,1240), NDWI(860,1640) and NDWI(1240,1640) are generated from the corn-oriented PROSAIL radiative transfer model. It is found that, instead of the linear relationship derived in previous studies, corn canopy water content (CWC) is best approximated as an exponential function of NDWI. Each NDWI variant is then applied to estimate corn CWC during the SMEX02, HiWATER2012, and Baoding 2018 field experiments. Both the simulation and validation results show that, among the four NDWI variants, NDWI(860,1640) and NDWI(1240,1640) have overall better estimation accuracies than NDWI(860,970) and NDWI(860,1240). When corn CWC is less than 1.0 kg/m2 or LAI is less than 2.0 m2/m2, NDWI(860,970) shows the best performance among the four NDWIs. However, when corn CWC is larger than 2.0 kg/m2 or LAI is larger than 5.0 m2/m2, corn CWC estimations from all four NDWI variants are unreliable. Based on the different performances of the four NDWI variants under different CWC and LAI conditions, an optimized NDWI-based scheme to estimate corn CWC is proposed. Under this newly optimized corn CWC estimating scheme, better CWC estimation accuracy is obtained - with the correlation versus ground truth (R) increasing from 0.81 to 0.87 and the root-mean-square error (RMSE) decreasing from 0.2326 kg/m2 to 0.1986 kg/m2. This optimized scheme is further applied to estimate corn VWC based on an observed linear relationship between corn CWC and vegetation water content (VWC) derived from in-situ corn CWC and VWC measurements. Compared with the estimation accuracy of the current SMAP-based corn VWC estimating method (R=0.74, RMSE=1.3304 kg/m2), our newly optimized scheme is superior (R=0.89 and an RMSE=0.6853 kg/m2).