Global gridded daily irrigation detection and quantification through soil moisture bias

Irrigation represents one of the primary anthropogenic perturbations to the global terrestrial water cycle, locally reshaping the climate-driven transition between wet and dry seasons through managed water inputs in croplands. Yet the globally consistent and temporally continuous daily irrigation estimates are still lacking. Here we interpret the persistent positive bias between remotely sensed and model-simulated soil moisture as an observable signature of irrigation and develop a global framework to quantify irrigation consumptive water use at the daily scale. We integrate multi-source inputs and construct a suite of representative scenarios to span major sources of uncertainty, improving robustness and internal consistency through observation-based constraints and fusion concepts. Independent consistency assessments and cross-region validation are further conducted to systematically evaluate the robustness, transferability, and uncertainty structure across gradients of climatic background and irrigation intensity. The global gridded daily irrigation figures more clearly delineate characteristic response patterns in major irrigated regions. In climate transition zones and strongly water-limited areas, estimates are more sensitive to climatic context and thus associated with relatively higher uncertainty. These findings provide a testable basis for interpreting regional differentiation and divergent magnitudes of irrigation impacts reported in the literature.