基于多项式回归模型的枯季径流预报与分析

Polynomial Regression Model Based on the Low Flow Forecasting and Analysis

冶河流域径流的补给来源主要是大气降水,径流和降水主要集中在5月-10月。由于受地质环境条件、流域下垫面因素作用,汛期径流与枯季径流存在良好的相关关系。枯季径流预报采用汛期径流量与枯季径流量建立多项式回归模型,通过对不同阶数的多项式进行对比分析,三阶多项式曲线拟合程度较好。采用符号检验和偏离数值检验法对该曲线进行检验,曲线拟合效果显著。并对枯季径流预报结果进行评价,按照枯季径流预报规范的要求,预报合格率达89.3%。该流域枯季径流,是沿河工业用水、灌溉用水的主要来源,分析和预报冶河枯季径流来水情况,为制定工农业用水方案提供科学依据。

The source of the Yehe River basin runoff is primarily originated from the precipitated water,mainly from May to October.Due to the effect of geological condition and basin underlying factor,there existed a positive relation between the flood season runoff and dry season runoff.The forecasting of the dry season runoff adopts a multinomial regression model.Through the analysis and comparison of the multinomial regression of different orders,the fitting curve of order 3 is preferable.Using the sign test and residual test method in the text of the curve,the fitting result is noteworthy.According to the demand of the dry season runoff prediction standard,the predicting result has a pass rate of 89.3%.The dry season runoff of this basin is the main source of water supply of industrial use and irrigation.Analysis and forecasting of the dry season runoff could provide scientific evidences to the scheme of water supply of agricultural and industrial use.