Prediction of evaporation from dam reservoirs under climate change using soft computing techniques

Area of Science:

• Environmental Science
• Water Resource Management
• Artificial Intelligence

Background:

• Accurate prediction of dam reservoir evaporation is vital for water resource management, especially under changing climatic conditions.
• Existing modeling techniques may not fully capture the complex interactions influencing evaporation rates.
• Climate change projections necessitate advanced tools to assess future impacts on water availability.

Purpose of the Study:

• To develop and evaluate artificial intelligence models for predicting evaporation from dam reservoirs.
• To assess the impact of climate change scenarios on reservoir inflow and evaporation.
• To compare the performance of different machine learning algorithms in modeling these hydrological processes.

Main Methods:

• Group Method of Data Handling (GMDH) and Least Squares Support Vector Regression (LS-SVR) were employed to model dam reservoir inflow.
• Adaptive Neuro-Fuzzy Inference System (ANFIS), optimized with Harris Hawks Optimization (HHO) and Arithmetic Optimization Algorithm (AOA), was used for evaporation modeling.
• IPCC's CMIP5 projections (RCP2.6, RCP4.5, RCP8.5) were utilized to predict future climate variables and their impact on the reservoir.

Main Results:

• LS-SVR demonstrated superior performance over GMDH in modeling reservoir inflow, achieving RMSE of 8.65 and NSE of 0.96.
• Hybrid models AOA-ANFIS and HHO-ANFIS significantly improved evaporation prediction accuracy compared to the standard ANFIS, with HHO-ANFIS achieving RMSE of 0.20 and NSE of 0.96.
• Climate change scenarios are projected to reduce inflow and increase evaporation, with impacts varying across RCP scenarios.

Conclusions:

• Hybrid machine learning models offer a robust approach for predicting dam reservoir evaporation and inflow.
• Effective dam reservoir management requires consideration of future climate change impacts on water resources.
• The developed AI models can be adapted for predicting evaporation in other reservoir systems.