资源环境科技发展态势分析平台

Regional climate models represent a valuable tool in climate impact analyses. Their ability to accurately estimate current and future climate conditions is increasingly important. In Norway precipitation is of special interest. Heavy precipitation, particularly over short durations, is responsible for enormous damages to infrastructure such as roads and railways, hence information on a fine spatial and temporal scale is crucial. We evaluate the ability of seven regional climate model simulations of 0.11 degrees resolution from the CORDEX ensemble in reproducing 3-h and 24-h accumulated summer precipitation characteristics in Norway. The two-step evaluation includes comparison of modelled precipitation to gridded observation-based datasets and to station measurements in terms of the following indices: summer maxima, summer wet event frequency, and total summer precipitation. We find a general overestimation by the models for all indices, with only few exceptions. Country-wide spatial averages show, however, that simulated summer maxima are mainly within the uncertainty interval of the gridded reference dataset. This might also be true for summer wet event frequency, although the comparison to station measurements indicates that the positive bias is significant. We find the largest deviation between models in the evaluation of summer totals. The spatial distribution of the different precipitation indices is fairly well simulated, although the precipitation gradients evident in observation-based datasets appear weak in the models. We believe the high spatial resolution improves the simulations of extreme precipitation in Norway, especially in areas of orographic enhancement.