Climate change impact on hydrological, sediment, and nutrient in Langat River basin: an assessment using integrated model / Nor Faiza Abd Rahman

Climate change, food security, water scarcity and environmental sustainability have all become major global challenges. In modelling the climate change impact, a growing issue on the lack of data management; cannot be solved simply by modelling simulation with missing data. There is an urgent need for innovation, for a better understanding of the current and future water resources, sediment and nutrient load in the river for water resources sustainability. Thus, the adoption of a strategic approach is necessary to planning and simulating the impact of climate change on hydrology, and its component for the respective authority can carry out its function and roles. This research aims are to study the infilling missing data techniques that are fast and reliable, and to speed up the weather data processing generation and impact of climatology on hydrology and its component that influence the development, planning, and management of successful semi-distributed climate assessment modelling in Selangor. The research suggested that artificial neural network (ANN) using a Lavenberg-Marquardt algorithm can successfully regenerate stream flow and sediment missing data for better accuracy of model. The results of the study also suggested that an automatic weather generator can simplify the preparation of weather data from six months to one month. A new automated weather generator input model for rainfall-runoff simulation (SWAT model) has been successfully developed to close the gap by integrating NCO, netCDF, Grads and CDO in a MATLAB environment. Regional daily weather variables were generated with XLs2Ncascii model that preserved the spatial and temporal dependencies and adequately reproduced statistics of the historic weather variables in the upper part of Langat River Basin. Nevertheless, the calibrated model provided an adequate measure of the effectiveness of Xls2Ncascii model coupling the weather generator with the watershed model and demonstrated a framework for streamflow, sediment and nutrient forecasting. The efficiency of the integration between Xls2Ncascii, ANN model output and SWAT model has proven to be improving in comparison using the SWAT model alone. R2 of 0.8 and NSE 0.75 thus prove that model integration is a great tool for prediction. The climate impact assessment shows that a non-consistent increase and decrease on streamflow with -100% to +250% impacts, sediment yield -100% to +2000% impact and nutrient analysis resulted in -100 to +800% impacts on climate change depending on climate change scenario, models and timespans. The findings of this study could contribute to the improvement of water management in Selangor to reduce the negative impact of climate change.