Timely monitoring of surface water is essential for water management in arid and semi-arid re gions. This study evaluates freely available Landsat imagery processed in Google Earth Engine to monitor water extent across f ive major Algerian reservoirs. Satellite-derived surface areas were compared with in-situ measurements from the National Ag ency for Dams and Transfers (ANBT), obtained from topographic and bathymetric surveys. Results show strong agreement bet ween satellite-derived and in-situ water surface areas, with correlation coefficients exceeding 0.97. Normalized errors, calcula ted relative to the normal reservoir surface area, remained low, with MAE ranging from 4.97% to 7.68% and RMSE from 5.70% to 8.37%. Linear regression models of the form Y = aX were developed for each reservoir, with slopes ranging from 1.0643 to 1.1093. These models enable reconstruction of historical surface areas and estimation of missing in-situ observations, especially before 2001. For ungauged reservoirs, the average slope provides a practical predictive equation. Leave-one-year-out cross-vali dation confirmed model robustness, while residual analyses showed no systematic bias. Overall, the approach provides a scalab le, cost-effective framework for operational reservoir monitoring, supporting near-real-time assessment, drought management, and long-term evaluation of reservoir dynamics under climate change, and facilitating evidence-based decisions for reservoir o peration, planning, and regional strategic water security.