A Proposed Downscaling Model for Climate Change Studies

摘要

General Circulation Models (GCMs) are commonly used to simulate future climate conditions in climate change studies. However the resolution from these models is too coarse for river basin scale studies. As such the results from these models need to be downscaled appropriately for use at the basin scale. Well known downscaling models include various dynamical and statistical approaches. However, none of these has been officially recommended as the best model for use in all regions. Disadvantages of existing downscaling models include high cost of operation, inability to avoid producing unrealistic values, inability to include multiple variables, and inability to reflect the future change of variability. In this paper, the development of a new downscaling model is based on a hybrid of algebraic and stochastic approaches that can include multiple variables is proposed. This new model will be called the HYAS model. HYAS employs the differences between simulated future and baseline variables added to simulated changing residual variance. This new approach has been applied in a climate change study of the Jangkok River Basin in Lombok, Indonesia. Twenty years (1971 to 1990) of GCM outputs were set as baseline variables and were used in the calibration process. The subsequent, twenty years (1991 to 2010) of data were used for model validation. Then, the relevant GCM outputs from 2011 to 2100 were used as model predictors to simulate the following regional climatic variables: humidity, rainfall, sunshine, temperature, and wind speed. Results showed that the following GCM variables: Screen 2m Temperature, Screen Specific Humidity, and Skin Temperature are appropriate for modeling regional humidity, rainfall, sunshine, and air temperature. However for modelling regional wind speed, the following GCM variables: Evaporation, Screen 2m Temperature, and Surface Pressure are more appropriate. The HYAS model is found to be superior to existing methods for predicting regional climatic variables in the Jangkok River Basin.