图形模式识别方法及其在中期雪灾天气预报中的应用

PATTERN RECOGNITION METHOD AND APPLICATION IN MEDIUM-RANGE FORECAST OF SNOW DISASTER

  • 摘要: 文章介绍了一种将计算机图形识别技术与数值预报产品相结合, 预报中期天气过程的图形模式识别方法和模式基元识别过程。 利用 ECM WF 的产品, 建立川西北高原雪灾中期预报模型。 在计算机中模仿预报员的思维步骤, 自动分类推理判别, 滚动预报未来 3~ 5 日内雪灾天气发生日期。 1998 ~ 1999 年雪灾中期预报效果显著。

     

    Abstract: Combined the computer pattern recognition and the numerical weather model products, the method of pattern recognition to predict the medium-range weather processes and the process of the pattern primitive element recognition are introduced. With the NWP products of ECMWF, the concept model of medium-range forecast of snow disasters in the northwestern plateau in Sichuan Province is established. The computer imitates the thinking process of forecasters, automatically classifies and judges the different categories to determine whether a snow disaster will occur in the future 3—5 days and which day it will occur. The results of the medium-range forecast of snow disasters are rather good.

     

/

返回文章
返回