预报因子及预报方程的优化方法探讨

Approach to the Optimizing Method of Selecting Forecasters and Prognostic Equation

造成中长期水文气象要素变化的因素多而且复杂，用单一的线性关系选择预报因子建立预报方程有可能把较好的预报因子漏掉。提出预报因子及预报方程的优化方法，采用多种非线性函数类型进行相关系数的计算。比较，确定优化预报团于建立优化预报方程，这样选择到的预报困子更符合实际，建立的预报方程具有更好的拟会预报效果。在预报方程建立过程中参数的计算是难点，使用二次插值和黄金分割法相结合计算参数α获得了成功。

There are many factors which cause the change of medium and long-term hydrometeorologic elernents. When the prognosticequation is Set up with forecasers selectal by single linear relationship, the better forecaser rnay be missed. The authors put forwardan optimizing method of selecting forecasters and prognostic equation by using various types of funtions to compute the correlation co-efficient.The forecaster eelected by this method will be more conformable to reality and the prognostic eqnation will be of fitter re-sults. In setting up the prognostic equation, paramter α is computed by interpolation in combination with golden section method.