基于水文模型与统计方法的中小河流致洪临界雨量分析

Analysis of critical flood causing rainfalls in medium and small rivers based on hydrological model and statistical method

以东津河为例,基于典型洪水过程的气象水文逐时观测资料,采用统计方法和水文模型建立了中小河流致洪临界雨量的分析方法流程,并进行了指标对比和效果检验。结果表明,东津河水位随雨量的变化与雨量累计时效明显相关,根据滑动累计相关分析结果确定16小时为临界雨量时效。通过回归分析建立了水位与累计面雨量的二次曲线关系,结合各等级洪水水位反算了对应不同前期水位的致洪临界雨量指标。与此同时,对TOPMODEL进行了模型率定和验证,经过率定后的TOPMODEL在东津河具备良好的适用性。基于水文模型和洪水水位-流量关系,采用试算法得到了另一套临界雨量指标。两种指标中,水文模型法相对而言更为敏感;实况检验表明,两种指标对洪水过程均有较好的指示意义,但预警等级与实况存在一定差异,统计方法倾向于低估,而水文模型指标则可能高估,综合两种指标在实际使用中应具有更好的预警效果。

Taking Dongjin River as an example, and based on temporal meteor-hydrological observation data on typ- ical flood process, analytical workflow of critical flood causing rainfalls in medium and small rivers was established, and index comparison and effect examination were carried out. Results showed that, the change of water level in Dongjin River was obviously correlated with rainfalls and the time efficiency of the accumulated rainfalls. Based on the moving accumulative correlation analysis, the critical rainfall time efficiency was determined as 16 hours. The statistical relationship between water level and accumulative rainfall was established by regression analysis. Com- bined with water level for each grade of flood, the flood-causing critical rainfall index values were calculated re- versely. At the same time, based on the calibrated hydrological model TOPMODEL and relationship between flood water level and flow rate, another set of critical rainfall index was obtained. Examination of actual condition indica-ted that the two sets of indices both have good applicability, but there are some difference between early warning grades and actual conditions. The statistical method tends to underestimate the actual condition and the hydrological model index may tend to overestimate the actual condition. Comprehensive application of the two indices should have good practical use effects.