Submitted to: Remote Sensing of Environment
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 3/13/2007
Publication Date: 2/15/2008
Citation: Yilmaz, M.T., Hunt, E.R., Goins, L.D., Ustin, S.L., Vanderbilt, V.C., Jackson, T.J. 2008. Vegetation water content during SMEX04 from ground data and Landsat 5 Thermatic Mapper Imagery. Remote Sensing of Environment. 112:350-362. Interpretive Summary: The Soil Moisture Experiment 2004 (SMEX04) was conducted in Arizona USA and Sonora Mexico during July and August 2004. One of the important factors that limit the ability of satellites to estimate soil moisture is the amount of water in vegetation. Liquid water in the leaves can be estimated with other satellites, such as the Landsat Thematic Mapper (TM). Reflectances of shortwave-infrared radiation at TM band 5 (1.55-1.75 'm wavelength) are reduced in relation to the volume of water per leaf area (equivalent water thickness), so vegetation indices with TM band 5 and TM band 4 (0.76-0.90 'm wavelength) can be used to estimate canopy equivalent water thickness. Field data from a wide range of vegetation types were measured during the SMEX04 experiment. Landsat TM imagery were acquired for three dates in 2004: June, July and August. Vegetation land cover for the SMEX04 experiment was determined by classification of the Landsat TM imagery. Vegetation indices using Landsat TM band 5 may be better for remote sensing canopy equivalent water thickness compared to an average value determined by land-cover class, because the vegetation indices are more sensitive to variation in leaf area index. However, to get total vegetation water content in the leaves, stems and roots from remote sensing, land-cover class must be used with the vegetation indices. These data will then contribution to better soil moisture retrievals from microwave satellites.
Technical Abstract: Estimation of vegetation water content is an important issue for soil moisture retrieval by microwave sensors and other remote sensing applications. The primary contribution of this study for the Soil Moisture Experiment 2004 was to sample vegetation for the Arizona and Sonora study areas. Determination of water in the foliage, measured by canopy equivalent water thickness (EWT), using reflectances in the shortwave infrared is a step towards the remote sensing of vegetation water content. Canopy EWT is defined as the EWT per leaf times the leaf area index; however, EWT per leaf and leaf area index are properties of the vegetation landcover. The normalized difference infrared index (NDII) was calculated from Landsat 5 Thematic Mapper bands 4 (0.76-0.90 'm wavelength) and band 5 (1.55-1.65 'm wavelength). NDII was linearly related to measured canopy EWT with an R2 of 0.54. Landcover of the Arizona and Sonora study areas was classified using a rule-based decision tree using three dates of Landsat 5 Thematic Mapper imagery and digital elevation data. The overall accuracy was only 70 %; however, the results were an indication that a distinct landcover class was not defined a priori. There was a large range of NDII per landcover class, perhaps due to variation in leaf area index, indicating that canopy EWT should be estimated directly using NDII.