10-meter Resolution Mapping of Farmland Shelterbelts in Northeast China (2019–2023)

This study employs multi-temporal Sentinel-2 imagery and leverages the unique phenological characteristics of shelterbelts to develop a multi-feature classification model that integrates spectral, texture, and temporal features. Using this approach, we successfully produced a 10-meter resolution farmland shelterbelt distribution map of Northeast China for the period 2019 to 2023. The results revealed a consistent year-on-year increase in shelterbelt length. The proposed method achieved an overall classification accuracy of 93.25% with a Kappa coefficient of 0.91. This study provides a novel approach for high-precision identification of farmland shelterbelts and offers valuable insights for remote sensing monitoring of other agroforestry ecosystems.