Innovative Crop Mapping in Hawaii: The First 10Meter Resolution Cropland Data Layers for 2023

This study developed the first 10-meter resolution Hawaiian Cropland Data Layers (HCDL) for 2023 using gap-filled Sentinel-Landsat multi-sensor 10-day composites and advanced algorithms. Evaluating both Random Forest and CNN models, the research found Random Forest to be more effective for operational purposes, achieving an average accuracy of $\mathbf{9 3. 6 \%}$. The HCDL provides a critical tool for Hawaii’s agricultural management, bridging data gaps and supporting informed decision-making.