This study presents an automated workflow for drought monitoring in Burdur Lake, Turkey, utilizing Sentinel-2 satellite data and K-means clustering. Five Sentinel-2 images from 2019 to 2023 were processed to derive spectral water indices. A water mask was generated by thresholding the indices, allowing for the distinction of water bodies. K-means clustering quantified changes in the lake area over time. The results reveal a decreasing trend in water extent from August 2019 to August 2023. In August 2019, the water extent was approximately 18.53%, which declined to around 16.64% by August 2023, signifying an approximately 10.3% reduction in water extent between the start and end years. This approach demonstrates a valuable framework for the integration of freely available satellite data and machine learning algorithms in operational drought monitoring.