Point to Area Rainfall Relations in the Context of Climate Change

摘要

For hydraulic structure design purposes, precipitation over an area for a specified duration and return period is commonly used in the estimation of floods. The main objective of the present paper is to propose a methodology for constructing the point-to-area rainfall relations in the context of climate change. In particular, a comparative study was carried out to assess the accuracy and reliability of the proposed method as compared to other existing methods using rainfall and climate data available from different sources in the southern Quebec and southern Alberta regions in Canada: observed daily rainfall data from rain gauges, Canadian Regional Climate Model (CRCM) output, and data given by different General Circulation Models (GCMs). The popular SDSM regression-based statistical downscaling method, the quantile-mapping, and the bias correction method were used to describe the linkage between large-scale climate variables given by the considered GCMs and local rainfall characteristics. Results of this illustrative application have indicated that the use of downscaling methods could provide the most accurate point-to-area rainfall relations as compared to the observed empirical ones, while the results given by the GCMs without downscaling and the CRCM were not accurate and displayed a very high level of uncertainty for both regions.