Climate change and increasing water demand threaten water security in strategic high-Andean reservoirs by altering seasonal water availability, increasing evaporative losses, and intensifying pressure on regulated storage. This study assesses the hydrological vulnerability of the Pañe Reservoir, Arequipa, Peru, under CMIP6 scenarios SSP2-4.5 and SSP5-8.5 and three water-demand trajectories projected to 2100. A reservoir-scale water-balance model was calibrated using a genetic algorithm with CHIRPS precipitation and ERA5-Land temperature data, achieving satisfactory performance in calibration (NSE = 0.886; $R^2$ = 0.907) and validation (NSE = 0.888; $R^2$ = 0.891). Climate projections from MPI-ESM1-2-LR, EC-Earth3-Veg-LR, and EC-Earth3 were bias-corrected using Quantile Delta Mapping. Under baseline demand, active storage maintains positive trends of +1.3 $\times {10}^6$ to +2.2 $\times {10}^6$ m³/decade; however, a 100% demand increase produces negative trends of up to −2.8 $\times {10}^6$ m³/decade under SSP2-4.5. Effective precipitation increases by +0.123 and +0.283 mm/day under SSP2-4.5 and SSP5-8.5, respectively, while Rx5day and Rx10day rise by +5.43/+8.78 and +12.13/+19.42 mm/decade. These trends suggest greater wet-season concentration of water inputs, but not necessarily improved dry-season security under rising demand. Reservoir warming reaches +0.60 °C/decade under SSP5-8.5, which may increase evaporative losses and could create conditions more favorable to water-quality deterioration. Overall, future vulnerability in the Pañe system reflects the imbalance between increasing withdrawals and the reservoir’s capacity to regulate highly seasonal high-Andean inflows, highlighting the need for demand control, loss reduction, bofedal protection and complementary wet-season storage.

Optimizing coagulant dosages in Drinking Water Treatment Plants (DWTPs) is critical for reducing operational costs, minimizing chemical waste, mitigating environmental impacts, and ensuring consistent water quality, particularly in resource-constrained settings where conventional jar tests are labor-intensive and poorly suited to real-time demands. This study develops and validates a Random Forest (RF) machine learning model to predict optimal dosages of aluminum sulfate, polyaluminum chloride, and a polymer flocculant at the Miguel de la Cuba Ibarra DWTP in Peru, addressing the need for an efficient, real-time decision support system. Using a historical dataset of 2556 jar tests, a univariate RF model was developed to predict settled water turbidity, tailored to the plant’s typical operational range. The model demonstrated robust predictive performance, achieving a coefficient of determination (R²) of 0.92 during training and 0.76 during validation with unseen data, alongside a Root Mean Square Error (RMSE) of 0.11 NTU and a Mean Absolute Percentage Error (MAPE) of 0.11 in the training phase. Integrated into a digital platform, the model generates real-time NTU ppm dosing curves, providing a practical and responsive tool to enhance operational efficiency for DWTP operators. This work offers a scalable, data-driven solution to improve water treatment processes in resource-limited contexts.