Development of a correction approach for future precipitation changes simulated by General Circulation Models

Producing reliable estimates of changes in precipitation at local- and regional-scales remains an important challenge in climate change science. Statistical downscaling methods are often utilised to bridge the gap between the coarse resolution of General Circulation Models (GCMs) and the higher-resolutions at which information is required by the majority of end users. However, the skill of GCM precipitation, particularly in simulating temporal variability, is not fully understood and statistical downscaling typically adopts a ‘Perfect-Prog’ (short for perfect prognosis) approach in which the derivation of high-resolution precipitation projections is based on real world statistical relationships between large-scale atmospheric ‘predictors’ and local-scale precipitation. Here, a ‘nudged’ simulation of the ECHAM5 GCM is conducted in which the large-scale climatic state is forced towards historical observations of large-scale circulation and temperature for the period 1958-2001. By comparing simulated and observed precipitation it is possible to, for the first time, quantify GCM skill in simulating temporal variability of precipitation. Correlation between simulated and observed monthly mean precipitation is shown to be as strong as 0.8-0.9 in many parts of Europe, North America and Australia. A nudged simulation permits the development of an alternative approach to statistical downscaling, known as Model Output Statistics (MOS), to correct precipitation as simulated by ECHAM5. It is also shown that MOS correction offers greater skill than Perfect-Prog methods when estimating local-scale monthly mean precipitation. The strongest-performing MOS models are applied to ECHAM5 climate change simulations and are shown to produce high-resolution precipitation projections that support those of RCM simulations. The potential for extending the MOS approach to daily precipitation is also assessed, with recommendations made for further research and application.EThOS - Electronic Theses Online ServiceGBUnited Kingdo