Improving agricultural water management in dryland mixed crop livestock systems using surface water bodies modelling and monitoring

Résumé fourni par la source

Abstract The mixed crop-livestock production in the dryland of Ethiopia is affected by the recurrent drought, necessitating the development of strategies for the sustainable water management and early warning system. The lack of real-time information on water availability at various waterpoints affects pastoralists travelling long distances in search of resources, frequently leading to competition and conflicts due to resource depletion, livestock loss, and increased vulnerability. Providing timely information on water availability is particularly challenging due to the absence of water monitoring and early warning system and scarcity of conventional station-based hydrometeorological networks in remote pastoral regions. This study advances water and drought management in mixed crop–livestock systems by (1) integrating citizen science for waterpoint water level monitoring, (2) calibrating and validating a hydrologic model to simulate waterpoint water level dynamics, (3) establishing a multi-source water balance model for continuous monitoring of water availability, and (4) introducing a novel operational envelope approach that translates simulated water levels into actionable operational classes for drought early warning and water management. Unlike previous water balance studies in dryland regions, this study integrates citizen science observations, waterpoint-specific hydrological calibration, and a novel operational envelope framework into a unified water monitoring and drought early warning system. By complementing conventional statistical validation with decision-oriented operational classification, the framework directly links hydrological simulations to water management actions and provides a foundation for future integration of seasonal and sub-seasonal climate forecasts. The calibration and validation of the water balance model showed satisfactory performance across all waterpoint systems with Nash–Sutcliffe Efficiency (NSE) and coefficient of determination (R 2 ) values of above 0.53 and 0.67, respectively, during the calibration period. Similarly, the model validation performed reasonably well at most waterpoints, except the Dingamo waterpoint. In addition to conventional statistical validation, this study introduces a novel operational envelope evaluation framework that assesses model performance based on the agreement between simulated and observed waterpoint operational classes (Good, Watch, Alert, and Near-dry). This operational validation directly links hydrological simulations to drought early warning and water management decisions. By demonstrating high agreement in operational classes, the proposed framework bridges the gap between hydrological modelling and decision support, thereby enhancing the practical value of the modelling framework for climate services and drought early warning. The early warning system can help provide vital water management advisory services tailored to pastoral and agro-pastoral communities.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.