科研之友
微信
新浪微博
Facebook
分享链接
摘要
Accurate and timely deterministic forecasts with long lead times are vital for the authorities in operational flood management decision making. Extending the lead time of streamflow forecasts can be achieved by using our one-way coupled hydrological and atmospheric modeling system (CHAMS). CHAMS has been used to generate real-time daily precipitation forecasts over China and subsequent flood forecasting in the Wangjiaba sub-basin of the Huaihe River Basin (HRB) since 2004. In this study, we first evaluate the skill of daily precipitation forecasts for the summers of 2005-2009 aiming at hydrological applications. We then examine the quality of daily flood forecasts for 10 major flood events encountered during the summers of 2005-2009. A total of 38 daily flood forecasts from the 10 events are examined. The result shows that CHAMS can well forecast both frequency and intensity distributions of the daily precipitation over the HRB especially in its central and northern regions. However, CHAMS generally overestimates the frequency of light to heavy rains, and underestimates the intensity of moderate to heavy rains. These uncertainties (or errors) in daily precipitation forecasts can have negative impacts on the subsequent streamflow forecasting. The 38 daily flood forecasts with different lead times compare reasonably well with observations. The average success rates of forecast peak discharges and peak times are 81.6% and 89.5%, respectively. Our result demonstrates that extending the lead time of streamflow forecasts beyond four days with adequate accuracy can be achieved by our one-way coupled modeling approach.
- 出版日期2013-7
- 单位河海大学; McGill
