Taking the Danba Section upstream of the main stream of the Dadu River Basin as an example,a comprehensive analysis of the water rise law in the basin was conducted.Combining the characteristics of runoff and the scheduling needs of downstream power stations,the threshold for water rise mode classification and the conditions for water rise judgment were determined using the traversal combination optimization method.A classified water rise runoff forecasting scheme based on"form-value"similarity was proposed,achieving a daily rolling forecast of runoff with a prediction period of 10 d.The results showed that from 2019 to 2021,the accuracy of identifying rising water in heavy rain,moderate rain,and light rain during the flood season in the Dadu River Basin reached 100％,93.18％,and 84.81％,respectively.Compared with the basic forecasting scheme that does not consider the identification of water rise conditions,the average relative error of runoff forecasting for different water rise modes is reduced by 0.45％to 6.81％for the classified water rise runoff forecasting scheme.The average relative errors of runoff forecasts with forecast periods of 1,5,and 10 d are 3.39％,9.64％,12.29％,respectively,and the Nash efficiency coefficients reach 0.99,0.95,and 0.91,respectively.This method has high forecast accuracy.
关键词
涨水识别/相似性预报/径流滚动预报/大渡河流域
Key words
water rise identification/similarity forecast/runoff rolling forecast/Dadu River Basin
维普
万方数据