📄 Abstract
Seasonal vegetation in the Charvak Reservoir basin (northeastern Uzbekistan) was evaluated using Landsat 9 Collection 2 Level-2 surface reflectance imagery for March, July, and October 2024. Cloud and cloud-shadow pixels were removed with the QA_PIXEL band. NDVI was calculated from SR_B4 (red) and SR_B5 (near-infrared) and summarized as monthly median composites. The three seasonal NDVI layers were stacked and classified in Google Earth Engine using unsupervised K-means clustering (k = 5; 5000 samples). Five spatially coherent classes with distinct seasonal NDVI signatures were obtained, separating persistently dense vegetation along valleys and moist slopes from seasonally variable hillslope vegetation, and from sparsely vegetated, bare-soil, and rocky surfaces on exposed ridges. The NDVI-cluster map provides a compact baseline for vegetation monitoring and spatial planning in the Charvak touristrecreational zone and highlights areas potentially sensitive to degradation under climate variability. Limitations include the single-year, three-snapshot design and the interpretive nature of unsupervised labels.
🏷️ Keywords
📚 How to Cite:
Dusmukhamedova I., Fazilova D.Sh., Aripova M.K , UNSUPERVISED MAPPING OF SEASONAL VEGETATION AROUND THE CHARVAK RESERVOIR , DOI: https://doi.org/10.36713/epra25338