俄罗斯模型中三个重要参数线型的统计分析

对俄罗斯模型中起决定作用的涨洪历时、洪峰流量和退水系数三个参数,在运用过程中均需通过线型优化技术、假设检验等确定线型,从而大大增加了模型运用过程中的工作量,并且容易出现抽样误差。为此,特收集了六个水文站的洪水资料,统计了三个参数的分布线型,通过比较假设检验、时段量模拟效果,发现三个参数多服从三参数对数正态分布。

Parameters Statistical Analysis of New Flood Stochastic Simulation Model

In the process of applying the new flood stochastic simulation model developed by Russian hydrologist Khristoforov, it always needs to use linear optimization techniques, hypothesis test to determine the line style of three important parameters in the model (rising flood duration, peak flow, regression coefficient), which greatly increase the workload and usually appear sampling error. Therefore, this study collects six hydrological stations data and estimates the line style of three parameters. It is concluded that three parameters can be supposed to obey three parameter lognormal distribution almost by using hypothesis test and time interval discharge statistical property test.