Temporal and spatial changes in grassland transpiration detected using Landsat TM and ETM+ imagery

Abstract

The water deficit index (WDI) derived from Landsat imagery was used to detect temporal and spatial changes in grassland transpiration. The WDI, which estimates relative evapotranspiration rates based on meteorological data and the relation between surface reflectance and temperature, has been successfully applied over heterogeneous terrain with little a priori information. In this study, WDI was derived from a 10-year, Landsat-4 thematic mapper (TM), Landsat-5 TM, and Landsat-7 enhanced thematic mapper plus (ETM+) data series of the Walnut Gulch Experimental Watershed in Arizona during the summer monsoon period. Our study showed that measurements of surface reflectance and temperature from the three sensors could be combined without sacrificing product accuracy. WDI was correlated (R2 = 0.73) with grassland transpiration measured by in situ instruments. Further, WDI varied temporally and spatially with variations in plant transpiration related to antecedent rainfall and slope aspect. WDI was compared with a measure of plant-available soil moisture (the antecedent retention index, ARI), which was derived from an hourly record of precipitation and runoff, obtained from rain gauges and flumes located in the watershed. Results showed that a nonlinear relation between WDI and ARI was significant but weak (R2 = 0.45) and implied that WDI was the more sensitive indicator of vegetation condition. Ultimately, the WDI approach may be used as a viable tool to monitor grassland condition over heterogeneous regions.