摘要

Urban hydrological climate change impact assessment requires an assessment of how short-term local precipitation extremes, e. g. expressed as intensity-duration-frequency (IDF) curves, are expected to change in the future. This cannot be done on the basis of regional climate model (RCM) results directly, mainly because of the models%26apos; coarse spatial grid and resolution-dependent limitations in process descriptions. In this paper, a stochastic model for downscaling short-term precipitation from RCM grid scale to local (i.e. point) scale is formulated and tested using data from Stockholm, Sweden. In principle, additional RCM variables characterising the weather situation are used to convert from grid box average to local precipitation, and estimate the probability of occurrence in an arbitrary point. Finally, local time series are stochastically simulated and IDF curves derived. The results of the application in Stockholm suggest that: (1) observed IDF-curves may be effectively reproduced by the model; and (2) the future increase of local-scale short-term precipitation extremes may be higher than the predicted increase of short-term extremes on the RCM grid scale.

  • 出版日期2012