遥测系统降雨观测粗差修正研究

Study on gross error correction of rainfall observation data using remote observation system

通过对遥测系统雨量观测资料中粗差特点和产生方式的分析 ,提出了借鉴曲面拟合算法的思路 ,利用抗差最小二乘估计和雨量空间分布函数修正粗差的方法。对闽江七里街流域 43个雨量站 46场暴雨 2 6万余条记录 ,用生成的误差系列 ,进行模拟估计 ,分析了此方法的有效性和合理性 ,获得了较好的结果。并且把修正过的降雨资料应用于新安江模型 ,对计算的流量和实测流量进行比较分析 ,达到了较满意的效果 ,表明本方法有较好的实用性。

In order to reduce the error of rainfall observation data at present, especially in the light of the situation of less research on the error of rainfall observation data using remote observation system, through the analysis on the gross error characteristics and occurrence mode of rainfall observation data by remote observation system, the authors suggest a gross error correction method by anti-error least square estimation and rainfall spatial distribution function through referring to curved surface fitting method. The simulation and estimation are carried out to more than 26 000 odd recorded data of 46 storms at 43 rainfall observation stations in Qilijie basin of Minjiang river system. The effectiveness and reasonableness analysis proves this method is satisfactory. The corrected rainfall data are applied in Xinanjiang hydrologic forecast model and the computed discharge and measured discharge are compared and analysed and the results are satisfactory, demonstrating the method is applicable.