HOW CHANGES IN HARVESTED AREAS OVER TIME AFFECTS THE ACTUAL EVAPOTRANSPIRATION OF CROPLANDS: THE CASE OF LOWER COLORADO RIVER

Advances in estimating actual evapotranspiration (ETa) with remote sensing (RS) have improved the accuracy of hydrological, agricultural, and climatological studies. RS-based ETa approaches provide spatial estimates that can be used across areas where in-situ data are scarce. The Vegetation Index-based ETa (ET-VI) method uses optical and near-infrared bands to calculate VIs and combine them with reference ET to estimate ETa. This study assesses the applicability of ET-VI method using Landsat-derived 2-band Enhanced Vegetation Index over croplands in the Lower Colorado River Basin, where croplands are highly dependent on irrigation and affected by drought events. We calculate ETa to monitor cropland water consumption and map agricultural drought (2000-2019). All calculations were conducted on Google Earth Engine platform. To compare the effect of harvested areas (HA) changes on the ETa rate we derived annual HA using the maximum Normalized Difference Vegetation Index time series. Shrinkage of crop area during dry periods can lead to ETa underestimation when changes in the HA are ignored (i.e., HA is assumed static for ETa prediction). Trend analyses of HA and ET-VI revealed that, although the extent of cultivated areas declined over time, ETa rates increased. This means that less HA did not lead to less water consumption of the region. This could be due to the greater cultivation of a water consumptive crop, i.e., alfalfa, one of the region's major crops. Our findings highlight the necessity of considering both HA and ETa rates in water management and drought monitoring of cropland as relying on one parameter can be misleading for decision-makers. It affirms the suitability of RS-based ETa as an efficient tool for water resources and cropland management. In future studies, we will evaluate the applicability of high-resolution sensors in the ETa derivation to monitor drought, water consumption, and to study the effect of crop patterns and HA on ETa estimation.