Attachments
Baseline assessment of land cover
A geospatial impact assessment was conducted to detect the significant negative changes of vegetation cover for agricultural land during the conflict of 2023-2024 in Lebanon. The assessment used the baseline period of October 2019 to 7th October 2023 compared to October 2024 without field validation. Land cover data is a critical resource for such geospatial analyses. The assessment began with the development of a 2023 land cover classification using Sentinel-2 imagery (10m resolution) and the national land cover legend. 1 137 training data for 18 land cover classes were gathered through a stratified random sampling approach.
Utilizing very high-resolution satellite Maxar imagery (<0.5m), crop calendars, and vegetation index profiles, the training data were prepared, and the Random Forest classification model was applied with Sentinel-2 satellite imagery. This land cover does not substitute national land cover dataset prepared by national mapping agency.