科研之友
微信
新浪微博
Facebook
分享链接
摘要
The use of Quantitative Precipitation Forecasts (QPFs) to model run-off and flow processes in urban areas is a challenging problem, as rainfall data with high spatial and temporal resolutions are required. Many attempts have been made to use weather radar to produce rainfall forecasts with lead times of a few hours ahead. %26lt;br%26gt;The UK Met Office in collaboration with the Australian Bureau of Meteorology has developed a stochastic probabilistic precipitation forecasting scheme (STEPS), which merges an extrapolation radar rainfall forecast with a high-resolution Numerical Weather Prediction (NWP) rainfall forecast. This paper assesses the application of this model in a small urban area (town of Yorkshire). Three precipitation events that occurred during 2007 and 2008 with different meteorological characteristics were simulated. The STEPS model was used to produce both deterministic and ensemble precipitation forecasts with spatial and temporal scales of 2 km and 15 min respectively and 6-hour lead time. The precipitation forecasts were coupled to an Infoworks CSmodel of the sewer system of a town in Yorkshire to produce flow predictions. The simulated precipitation events were analyzed in terms of rainfall and flow predictions at the urban scale of the study area. %26lt;br%26gt;The results show that the overall performance of the rainfall forecasting system decreases with increasing rainfall intensities, and that the ensemble rainfall forecasts have a higher skill than the deterministic forecasts in predicting lower rainfall intensities. The results also show that stratiform precipitation is forecasted better than convective precipitation. %26lt;br%26gt;More events need to be evaluated in order to define whether ensemble rainfall forecasts improve flow predictions on the urban scale and the analysis in terms of flow at this stage only supports a potential application for qualitative flood warnings in the small urban catchment considered.
- 出版日期2012-1
