Hyper-resolution modeling of crop evaporation in a semi-arid region using GLEAM and METRIC

Water scarcity is a major challenge for effective agricultural water management in semi-arid regions. The lack of water resources often requires irrigation (e.g., surface and sprinkler irrigation), providing crops with sufficient soil moisture to maintain photosynthesis and transpiration. To improve crop yields and simultaneously minimise water usage, accurate monitoring of crop evaporation, the primary indicator of plant water consumption, is essential. Given the heterogeneity inherent to semi-arid croplands, hyper-resolution images can enhance the quality and accuracy of monitoring(<30m). Such monitoring systems necessitate the development of remote sensing-based models capable of resolving processes at hyper-resolution and providing spatio-temporally consistent estimates of evaporation.
In this study, we estimate daily crop evaporation of wheat in the experimental site of the Agriculture College of Shiraz University (Shiraz, Iran) over four years (2016–2020). As a first step, we drive the Global Land Evaporation Amsterdam Model (GLEAM) with Landsat 8 data to generate evaporation at hyper-resolution (30 m). The GLEAM model, originally designed to estimate evaporation at ecosystem-to-global scales, is adapted to consider both surface and sprinkler irrigation in water balance calculation, a common feature in irrigated agriculture. The additional water through surface irrigation is introduced into the system via the soil water balance module, whereas the sprinkler irrigation is introduced as additional precipitation into the interception module. In a second step, we execute an energy balance model, the Mapping EvapoTranspiration at high Resolution with Internalized Calibration (METRIC), using Landsat 8 data. When appropriate extreme pixels (hot and cold pixels) are specified, METRIC can calculate advection, and also METRIC performance is accurate under heterogeneous land use. The results of these two distinct approaches are intercompared and validated against in situ data.