摘要
以西藏那曲地区为研究目标,采用多元线性回归法对多个预报因子与预报量之间的关系进行了分析,建立了预报量与预报因子之间的多元线性回归方程,并依据对强降水天气个例物理量统计所得出的预报指标,利用回归方程来对未来某天强降水天气的降水量进行预报。结果表明,西藏那曲地区未来24 h降水量的预报方程为y=16.861-0.185x1+0.164x2+0.234x3,且回归方程通过显著性检验;利用回归方程对检验个例的误差统计发现,相对误差较小,预报结果可以接受。
With Tibet Naqu area as the research target, multiple linear regression method was adopted to analyze relationship between multiple fore- cast factors and forecast quantity, the multiple linear regression equation was established. According to forecast indicators obtained from the case of heavy rain weather parameters statistics, the regression equation was adopted for forecasting heavy rain in the future. The results showed that the fore- cast equation for future 24 h precipitation in Tibet Naqu area is y = 16, 861 -0. 185x1 +0. 164x2+0. 234x3, and the regression equation passed the sig- nificant test; through error statistics, it was found that the ralative error is smaller, the forecasting results could be accepted.
出处
《安徽农业科学》
CAS
2013年第13期5855-5856,5861,共3页
Journal of Anhui Agricultural Sciences