This study proposes a digital twin-based model for calculating crop water requirements in irrigation districts, establishing a closed-loop management system from data acquisition to decision support. The model adopts a four-layer progressive architecture: the foundational data layer integrates multi-source remote sensing, DEM, and land use data; the feature information layer extracts key parameters including crop distribution, soil properties, and meteorological factors; the calculation layer combines water balance principles with the Penman–Monteith equation for accurate estimation; and the visualization layer employs WebGIS for interactive results display through heatmaps and animations. Validation in the Tingzikou irrigation district demonstrates the model's capability to dynamically track spatiotemporal water demand patterns, providing scientific support for optimized irrigation scheduling. The framework offers an innovative and practical digital twin solution for precision agricultural water management.