Skip to main content
ARS Home » Research » Publications at this Location » Publication #213744
Title: Remote sensing of vegetation water content from equivalent water thickness using satellite imagery.

Author
item YILMAZ, M. TUGRUL - GEORGE MASON UNIV.
item Hunt Jr, Earle
item Jackson, Thomas

Submitted to: Remote Sensing of Environment
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/27/2007
Publication Date: 5/15/2008
Citation: Yilmaz, M.T., Hunt, E.R., Jackson, T.J. 2008. Remote sensing of vegetation ater content from equivalent water thickness using satellite imagery. Remote Sensing of Environment. 112:2514-2522.

Interpretive Summary: Soil moisture content can be estimated from active or passive microwave remote sensing. However microwaves are also affected by the amount of water in vegetation, which lowers the accuracy of the estimated soil moisture content. If vegetation water content can be remotely sensed with other sensors, then the accuracy of estimated soil moisture content will increase. Water in the foliage strongly absorbs shortwave infrared radiation at about 1600 nm wavelength. An index called the normalized difference infrared index (NDII) based on the difference in leaf reflectance at 850 nm and 1600 nm is linearly related to the leaf water content per leaf area. The Soil Moisture Experiment 2005 (SMEX05) was conducted near Ames, Iowa in June and July 2005 to validate methods for the remote sensing of soil moisture content. We collected data on leaf water content and total water content for corn, soybean and deciduous hardwood woodlands and found that the NDII was linearly related to the water content of the vegetation canopy. Furthermore, the total vegetation water content was linearly related to the canopy water content for corn and soybean, but not the deciduous hardwoods. Allometric relationships are based on a comparison between the size or mass of one part of a plant with the size or mass of another part of the plant. There is an allometric relationship between the canopy water content and total vegetation water content, which allows the estimation of total vegetation water content from NDII. Land cover classification of remotely sensed images enables the use of different allometric relationships for different crop types or ecosystems. Therefore, total vegetation water content can be remotely sensed using shortwave infrared reflectances, but it depends on stable allometric relationships with the canopy water content.

Technical Abstract: Vegetation water content (VWC) is one of the most important parameters for the successful retrieval of soil moisture content from passive and active microwave data. Normalized Difference Infrared Index (NDII) is a widely-used index to remotely sense Equivalent Water Thickness (EWT) of leaves and canopies, but the amount of water in the foliage is a small part of total VWC. In this study, NDII and EWT were related to VWC for different landcover classes using multispectral satellite imagery. Vegetation data sampled during the Soil Moisture Experiment 2005 (SMEX05) in the Walnut Creek watershed near Ames, Iowa. Sites of corn (Zea mays), soybean (Glycine max) and deciduous hardwood woodlands were sampled to estimate EWT and VWC. Using a time series of Landsat, ASTER and AWiFS imagery, NDII was related to EWT with R2 of 0.85. Also EWT was linearly related to VWC with R2 of 0.88 for corn and 0.47 for soybeans; however, there was no significant relationship found between EWT and VWC for deciduous hardwood woodlands (P = 0.81). The 2005 land cover classification product from the USDA National Agricultural Statistics Service had an overall accuracy of 92% and was used to spatially distribute VWC over the landscape. SMEX05 VWC versus NDII regressions were compared with the regressions from 2002, which were conducted in the same study area. No significant difference was found between years for corn (P = 0.13); whereas there was a significant difference for soybean (P = 0.04). Stable allometric relationships between EWT and VWC are required to estimate VWC from NDII.