Rule-based 10 m resolution mapping of winter wheat in the Nile Delta based on growth rate difference under fuzzy phenology

Egypt’s wheat production capacity remains insufficient to achieve self-sufficiency, creating long-term reliance on imports to safeguard food security. Accurate identification and mapping of winter wheat are thus critical for effective agricultural management and grain supply monitoring. However, most existing mapping products focus on major wheat-producing regions, providing limited coverage for countries such as Egypt. The challenge is further compounded by Egypt’s unique intercropping of winter wheat and clover, which complicates discrimination between the two crops. To address this, we propose a rule-based, sample-light winter wheat identification method based on temporal differentiation of surface features. First, vegetation and non-vegetation were separated; then fuzzy phenological features were extracted by exploiting growth rate differences between winter wheat and clover. Using this approach, we generated annual 10 m resolution winter wheat distribution maps for the Nile Delta from 2020 to 2024. The mapped areas showed strong agreement with official statistics, with determination coefficients (R2) consistently above 0.844. In the XGBoost-based feature-benchmarking experiment, the proposed NDVI-NMILNE-GRDI feature combination outperformed conventional spectral, texture, SAR backscatter, and vegetation-index features, achieving an overall accuracy of 0.949. This methodology enables rule-based, sample-light winter wheat mapping with reduced dependence on extensive ground-based sampling, offering substantial potential for application in regions where detailed field observations are limited.