Electronic Thesis and Dissertation Repository


Master of Engineering Science


Civil and Environmental Engineering


Dr. Slobodan P. Simonovic, Dr. A. Ian McLeod


Climate change has the potential to significantly alter the hydrologic cycle, changing the frequency and intensity of precipitation events in an area. It is necessary to quantify these effects to effectively manage water resources in the future. Atmosphere-Ocean coupled Global Circulation Models (AOGCMs), often used in climate change research, have spatial resolutions that are too large to capture the local climate characteristics of a watershed. As a result, several downscaling tools have been developed, including stochastic weather generators. A methodology for the simulation of historical and future climate data using a nonparametric K-Nearest Neighbour block resampling weather generator with perturbation is presented (KnnCAD Version 4). The proposed approach is illustrated using a case study of the Upper Thames River basin in Ontario, Canada. KnnCAD V4 is shown to effectively reproduce the historical climate and can produce future climate change scenarios based on AOGCM data.