Evaluation of Variable-Infiltration Capacity Model and MODIS-Terra Satellite-Derived Grid-Scale Evapotranspiration Estimates in a River Basin with Tropical Monsoon-Type Climatology

With the limited availability of meteorological variables in many remote areas, estimation of evapotranspiration (ET) at different spatiotemporal scales for efficient irrigation water management and hydrometeorological studies is becoming a challenging task. Hence, in this study, indirect ET estimation methods, such as moderate resolution imaging spectroradiometer (MODIS) satellite-based remote-sensing techniques and the water-budget approach built into the semidistributed variable infiltration capacity (VIC-3L) land-surface model are evaluated using the Penman-Monteith (PM) equation approach suggested in the literature together with a crop coefficient approach. To answer the research question of whether regional or local controls of a river basin with tropical monsoon-type climatology affect the accuracy of the VIC and MODIS-based ET estimates, these methodologies are applied in the Kangsabati River Basin in eastern India at 25×25 km resolutions attributed with dominant paddy land uses. The results reveal that the VIC-estimated ET values are reasonably matched with the PM-based ET estimates with the Nash-Sutcliffe efficiency (NSE) of 54.14–71.94%; however, the corresponding MODIS-ET values are highly underestimated with a periodic shift that may be attributed to the cloud cover and leaf shadowing effects. To enhance the field applicability of the satellite-based MODIS-ET products, these estimates are standardized by using a genetic-algorithm-based transformation that improves the NSE from −390.83 to 99.57%. Hence, this study reveals that there is the need of a regional-scale standardization of the MODIS-ET products using the PM or lysimeter data or possible modification of the MOD16A2 algorithm built-into the MODIS for generalization. Conversely, the satisfactory grid-scale ET estimates by the VIC model show that this model could be reliably used for the world’s river basins; however, at smaller temporal scales, the estimates could be slightly inconsistent.