Abstract
AbstractClimate change impact reports by the Intergovernmental Panel on Climate Change (IPCC) are well established throughout the world. By now many local, regional, and central authorities are aware of these impacts and they try to take necessary precautions according to their environmental circumstances. Preliminary decisions cannot be taken unless there is a model for future climate change impacts, especially for rainfall occurrence, amount, and frequency estimations. The main purpose is to combine the general circulation (climate) model (GCM) outputs with the local meteorology station records by a suitable downscaling procedure. In this paper, a statistical quadrangle downscaling model (QDM) is developed for rainfall estimations up to 2100. This model helps to downscale the nearby four scenario rainfall amounts at GCM grids to a desired point. Two main steps in the model structure are the use of a regional dependence function (RDF) for transferring the GCM output data from the four nearest grid points to the location. The second step is to adjust the downscaled scenario monthly rainfalls with the local rainfall records. The GCM monthly rainfall outputs are downloaded from the National Center for Atmospheric Research (NCAR) with special report emission scenario A2. The application of the methodology is presented for Riyadh City, Kingdom of Saudi Arabia (KSA). The RDF is obtained from the most reliable meteorology records at 27 locations that are irregularly scattered all over the country. Then it is used for scenario data transfer from the regular grid points. The adjustment is based on the average statistical properties of concurrent monthly rainfall amounts between 1985 and 1996. The trend component in the GCM remains intact. It has been observed that there is a major increasing trend in the monthly rainfalls after 2050. In the meantime, each decade starting from 2011 gives estimates that indicate steady rainfall increases up to 2050. The QDM indicates a 40% increase in the annual total rainfall amounts, with a 15% standard deviation increase.