AquaLeach approach: Field-scale irrigation assessment using satellite-derived products while accounting for salinity leaching fraction in arid and semi-arid irrigation schemes of Kenya

Irrigation water management in arid and semi-arid regions is increasingly challenged by water scarcity and soil salinization. Quantifying how irrigation water is partitioned between crop use and leaching remains critical but is often hindered by the lack of field measurements. This study presents the AquaLeach model, a satellite-based framework that integrates the SM-based inversion approach for irrigation estimation with a leaching fraction (LF) module to assess the portion of applied water contributing to salt leaching. The model was implemented over two large irrigation schemes in Kenya, Perkerra and Tana, using multi-year (2018–2023) remote-sensing data from the FAO WaPOR platform. Benchmark irrigation records and reference leaching requirement (LR) values derived from soil and water salinity data were used for validation purposes. AquaLeach reproduced the magnitude of irrigation events, with average percentage errors of −9.3% in Perkerra and −18.6% in Tana. Mean LF values ranged from 0.12 to 0.20 in Perkerra and 0.33 to 0.49 in Tana, showing consistent temporal variability with LR-based reference estimates (range 0.03–0.10) despite differences in absolute magnitude. These results demonstrate the capacity of AquaLeach to provide field-scale, multi-season assessments of irrigation performance and salinity control using only satellite-based inputs. The framework offers a scalable and transferable tool for advancing irrigation monitoring and supporting climate-resilient agricultural water management in salt-affected environments. • AquaLeach introduces a unified satellite-based framework to jointly estimate irrigation and leaching fraction (LF). • The SM-based inversion reproduced multi-season irrigation dynamics in Perkerra and Tana, with annual errors within ± 30%. • LF estimates captured spatial heterogeneity and interannual variations consistent with independent LR benchmarks. • The framework provides a scalable, data-sparse solution for irrigation performance and salinity-risk assessment.