COMPARATIVE ASSESSMENT OF VEGETATION CONDITION IN AGRICULTURAL LANDS OF FERGANA REGION USING SENTINEL‑2 AND LANDSAT 8–9 NDVI

Abstract

This paper compares vegetation condition in irrigated agricultural lands of the Fergana Region (Uzbekistan) using NDVI derived from Sentinel‑2 MSI and Landsat 8–9 OLI imagery. NDVI maps were generated for June 2025, a key period of crop development in the valley. A unified processing workflow was implemented in ArcGIS Pro: scene screening for low cloud cover, radiometric conversion to reflectance (using Level‑2 products where available), cloud/invalid‑pixel masking, band compositing, and clipping to the regional boundary. The resulting NDVI rasters were compared visually and through simple spatial diagnostics to evaluate (i) the clarity of field boundaries, (ii) within‑field heterogeneity, and (iii) consistency of high/low vegetation zones. Both sensors reproduce the regional pattern of vegetation vigor; however, Sentinel‑2 provides substantially clearer field‑scale detail due to its 10 m red and NIR bands, while Landsat 8–9 (30 m) shows smoother NDVI surfaces and is more suitable for retrospective and long‑term analyses because of its long archive. The study confirms that a multi‑sensor approach can improve operational agricultural monitoring in fragmented irrigated landscapes.