Multidecadal Trend of Basin‐Scale Evapotranspiration Estimated Using AVHRR Data in the Krishna River Basin, India

Abstract

Evapotranspiration (ET) is one of the most critical components of terrestrial water balance. Estimating reliable ET across a large region is, however, difficult due to the limited number of ground monitoring stations and the heterogeneous land surface conditions. In this work, spatially distributed monthly ET estimates in the Krishna River Basin, India, were derived using Advanced Very High Resolution Radiometer (AVHRR) data and the ground monitoring data sets from local meteorological stations for the 1983 to 2001 time period. A modified Penman–Monteith model with biome‐specific canopy conductance was used to estimate the 8‐km AVHRR ET. Ground validation was conducted using the lysimeter measurements and Landsat‐ and MODIS‐based ET estimates from the same model aggregated to the AVHRR scale. The difference between the lysimeter measurements and 30‐m Landsat estimates was only 5.3%, and the mean difference between the aggregated Landsat ET and the AVHRR ET estimates was <14%. The AVHRR ET estimates of this study showed that the basin‐average ET increased continuously at the mean rate of 4.97 mm yr−1 yr−1 in the Krishna River Basin during 1983 to 2001. The subbasin‐scale ET analyses showed an increasing trend of ET in the upstream subbasins, where irrigation development was active in the 1980s, compared with the other subbasins. Further sensitivity analyses based on land‐use information showed that the increased ET in the basin is mainly due to the development of irrigation that occurred during the study period.