Statistically downscaling of climate change

• Main Author: Tang, Xin Ning
• Other Authors: Qin Xiaosheng
• Format: Final Year Project
• Language: English
• Published: 2016
• Subjects: DRNTU::Engineering
• Online Access: http://hdl.handle.net/10356/67249
• Institution: Nanyang Technological University

As global warming and increasing emission of greenhouse gases have gained much concern from scientists around the world, studies on climate impact is important. Global Climate Models (GCMs) is the most advanced technology and models that are widely used to facilitate their study on climate change. However, GCMs usually have a coarse spatial resolution which is ill in providing accurate regional climate change information. In order to overcome this obstacles, scientists have introduced numerous downscaling techniques to refine the coarse resolution GCMs to obtain local climate information. Amongst the downscaling techniques, a multiple regression-based model namely Statistical Downscaling Model (SDSM) has gained its popularity around the world in downscaling weather series. In this study, SDSM was applied to downscale rainfall and temperature at Singapore Changi Airport, provided observed weather data from weather station. The study included the evaluation of model calibration and validation in SDSM with National Centers for Environmental Prediction (NCEP) re-analysis predictor variables. Then, predictions of future rainfall and temperature in SDSM with CGCM3 predictors corresponding to environment scenario A2 were carried out. The study results shows that SDSM is capable in model calibration and validation stages. However, it is relatively incompetent in projection for future weather series especially for conditional models like rainfall. There is a significant increment or decrement trend for several months.