基于降雨分级的泥石流降雨I-D预报模型

Study of I-D Relationship for Debris Flows Based on Rainfall Classification

降雨阈值是泥石流预警预报和防灾减灾的重要研究内容。通过野外考察、文献查询等手段,收集了汶川地震灾区典型区域(都(江堰)汶(川)公路沿线)震后34次泥石流事件和对应的降雨过程,通过统计诱发泥石流的降雨历时及平均降雨强度,建立了该地区的降雨I-D(雨强-历时)关系;针对I-D关系在实际预报中的不足,结合中国气象部门的雨量划分标准和诱发泥石流的降雨历时将降雨分为小量级降雨(SI)、中量级降雨(MI)和大量级降雨(LI)3个等级,得出复合型降雨I-D关系。结果表明:①都汶地震灾区诱发泥石流的降雨历时为2~53h,平均降雨强度为0.1~13.8mm/h,I-D关系可表示为I=5.94 D-0.70;②复合型降雨I-D预报模型可进行分级预警,其中约80%泥石流事件超过了橙色等级预警线,I-D关系式为I=10.0 D-0.56(1h≤D<12h),I=4.8 D-0.26(D≥12h);③改进后的复合型降雨I-D阈值线斜率更小,预报精度更高,更适合中国西部山区的泥石流预报工作,且在2014-2016年都汶公路沿线的泥石流预报检验中取得了较好效果。本研究拟为汶川地震灾区泥石流的预警预报提供技术支撑。

Identification of rainfall thresholds is crucial for debris flow forecasting and hazards mitigation.The data of 34 earthquake debris flow events and their responding rainfall processes in a typical debris flow prone region(the zone along the Dujiangyan-Wenchuan National Highway)were collected through field investigations,interviews,and literature researches,etc.The I-Drelationship is established by analysis of the rainfall duration and mean rainfall intensity.A new debris flow warning classification is proposed based on the rainfall division method of the meteorological department of China and the rainfall duration of the debris flows.The rainfall events are divided into three levels:small-scale rainfall(SI),medium-level rainfall(MI),and large-level rainfall(LI),and the I-D thresholds are improved.The results indicate:(1)The rainfall duration ranged from 2 to 53 hours;the mean rainfall intensity was in the range of 0.1 to13.8 mm/h;and the I-D relationship can be expressed as I=5.94 D-0.70.(2)The compounding I-D prediction model can be used for hierarchical warning,and about 80%debris flow events exceed the orange warning line,which can be expressed as I=10.0 D-0.56(1 h≤D<12 h)and I=4.8 D-0.26(D≥12 h).(3)The improved compounding I-D relationship has a smaller slope and higher accuracy,and more suitable for the debris flow forecasting in the mountainous areas of the western China.The results are tested by the debris flows in the period of 2014 to 2016,which shows an acceptable accuracy.This study is intended to improve debris flow forecasting and hazards mitigation for earthquake-prone regions.