A process-based modelling of groundwater recharge under contrasting irrigation methods in semi-arid crops

We present an application of the process-based ecohydrological model EcH2O to evaluate water-energy coupling and resulting percolation beneath the root zone under contrasting irrigation practices in a semi-arid region. The study uses high-resolution data from two wheat fields employing flood and drip irrigation, in a multi-objective calibration and evaluation approach with datasets encompassing soil water content at two depth ranges, energy balance components, and percolation rates at two depths. We find that the model reasonably simulates water fluxes and energy partitioning, and captures the distinct hydrological responses of the different irrigation methods. The best overall performances were found at both sites using calibration scenarios combining all available datasets, pointing at complementary information footprints. These footprints were nonetheless heteregeneous, as for example simulation of energy balance components showed little change between calibration scenarios, while percolation fluxes were acceptably captured only if the corresponding datasets were included in the calibration. Results highlight larger percolation dynamics and amounts beneath the root zone of flood-irrigated wheat, yet the two indices used here for irrigation efficiency reveal opposite rankings between the two irrigation methods depending on whether deep percolation is included (as a proxy for aquifer recharge) or not in the hydrological system being analysed. These findings challenge the view on greater water-saving benefits associated with drip irrigation, given the complex trade-offs between irrigation amounts and timing, plant water use, and return flows (e.g. underlying aquifer recharge). This analysis is a step forward for informing integrative and sustainable water management strategies in arid and semi-arid agricultural contexts.