Integrating biophysical models and remote sensing to evaluate irrigation practices in four global hubs

Sustainable agricultural intensification through irrigation is necessary to address the challenges that climatic and demographic change pose to the global food system. This requires creating accurate regional-to-global knowledge bases of irrigation estimates, in a context where the availability of ground data is as valuable as infrequent. Large scale agro-hydrological models and irrigation estimates from earth observations are often associated with underrated yet significant uncertainties, but they also have complementary strengths and weaknesses. While already used in synergy for field-to-district scale applications, their synergistic use at larger scales remains unexplored. To fill this gap, we present a novel comparison of irrigation demand simulations from a spatially distributed agro-hydrological model and irrigation water use estimates from five satellite retrievals, over four global irrigation hubs. Despite describing different variables with independent tools running on independent data, the results show statistically significant linear correlations above 0.6 between biophysical simulations and satellite retrievals for three cases out of four. Moreover, space-time discrepancies pinpoint irrigation responses to hydroclimatic and anthropogenic drivers. Thus, a synergistic use of earth observation and large-scale agro-hydrological modelling, beyond mere data input, can improve our understanding of coupled human-natural dynamics in irrigation. ● We compare biophysical water demand with water use from earth observation in four global irrigation hubs. ● Despite different inputs, methods, and non-identical simulated variables, we find good agreement across estimates. ● Space-time discrepancies between signals reveal irrigation water use responses to hydrometeorological fluctuations.