Future reservoir performance projection using integrated models in the framework of climate change
Loading...
Date
2016
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Universiti Teknologi Malaysia
Abstract
Climate change has made reservoir system management ever more challenging due to a general lack of contingency plan to handle drastic water supply and demand changes in the optimal operation for a sustainable and long term reservoir system planning. The main objective of this study is to propose a long term optimal reservoir operation on water release which considers climate changes for the Bukit Merah Reservoir. The proposed integrated models contained a climate model, a hydrological model and an optimization model, and were tested for two climate emission scenarios, i.e., the Global Climate Models (GCMs) simulations from Hadley Centre 3rd generation with scenario A2 (HadCM3 A2) and A1B. The A2 scenario had Statistical Downscaling Model (SDSM), Artificial Neural Network (ANN) model, and Linear Programming (LP) and Particle Swarm optimization (PSO) models. The A1B scenario had Providing Regional Climates for Impacts Studies (PRECIS) results obtained from the Malaysian Meteorological Department (MMD), ANN model, and LP and PSO models. Both scenarios were used to develop the optimal water release for Bukit Merah reservoir. Three performance measuring indices - reliability, resiliency, vulnerability were gauged against historical reservoir performance. All simulations reported hereafter were for the next 100 years. The first integrated model using SDSM model showed an increased in temperature (0.5 to 2.2 oC) and rainfall (35%). Paddy irrigation demand will reduce 15% per century due to increase of rainfall. Flow simulation using ANN models from SDSM downscales showed an increase in mean daily flow (17%). The PSO and LP models successfully met the objective function and established a sustainable long term reservoir operation. The second integrated model from PRECIS showed an increase in temperature (1 to 3 oC) and rainfall (12%). Paddy irrigation demand will reduce 11% per century. Flow simulation using ANN model show an increase in mean daily flow (15%). The PSO model performed better than LP model. In term of reliability, resiliency and vulnerability, PSO model shows sufficient water supply in the future. In conclusion, this study has successfully contributed to the development of models using evolution algorithm, hydrologic model and statistical methods for sustainable water resources planning and management in consideration of future climate changes.
Description
Thesis (PhD (Civil Engineering))
Keywords
Reservoirs—Management, Reservoir drawdown