Tile-drainage and Crop Rotation Enhanced Cropland Dataset to Improve Spatial Accuracy of Eco-hydrologic Models

Ecohydrological modeling has advanced through integration of geospatial datasets, where input data accuracy and spatial resolution critically influence model representativeness and estimates of water balance and streamflow. In the U.S. Corn Belt, accurate row-crop systems and tile-drainage representation can significantly impact the water-quality simulation. Existing datasets lack collective classification on crop-rotations and tile-drainage, limiting their spatial representativeness. We developed the 30m-resolution "Tile-drainage and Rotation-Enhanced Cropland" (TREC) dataset for Conterminous United States, integrating both continuous-crop rotations and tile-drainage data. We assessed the performance of TREC against the traditional Cropland Data Layer (CDL) using SWAT + model in two Midwestern watersheds. Simulated tile-flow using TREC remained comparable to CDL (147.6 vs. 147.4 mm) in uncalibrated simulations, while TREC produced more localized tile-drainage consistent with mapped tile-drainage patterns from AgTile-US, improving spatial representation compared with CDL. Across monitoring stations, TREC (KGE = 0.54-0.6) improved hydrologic performance relative to CDL (KGE = 0.25-0.59), reflecting the structural benefits independent of calibration effects. This enhanced precision supports reliable hydrologic-modeling and facilitates targeted decision-making for watershed conservation, nutrient management, and agricultural-policy development.