基于模式预报倾向的预报系统设计与初步试验

Forecast System Design and Preliminary Tests Based on Model Forecast Tendency

为了充分利用数值模式的预报产品信息,首先对数值模式预报产品的误差进行了分析,针对误差分析结果,首次提出了一种新的基于模式预报倾向的精细化预报方法,并在此基础上设计开发了新疆精细化数值预报系统。初步试验结果表明:该方法对新疆地区的温度预报具有一定的正效果。后期经过调整订正系数,利用更多时次的观测资料进行温度订正,则有望可以使订正的温度更加接近观测。即该方法可以提高气象台站的温度预报准确率,展现了该方法的业务应用潜力。

To fully use the information of forecast products from numerical models, the errors of the prod- ucts are analyzed. According to the results of the analysis, a refined forecast method is proposed and based on it a new Xinjiang refined numerical forecast system is developed. Preliminary tests are conducted and the results show that the new method has a positive effect on temperature forecasting in the area of Xinjiang. After adjustment of correction coefficients and usage of more observations at different time points, the corrected temperature is expected to be much closer to the observations. That is, the method can improve the accuracy of temperature forecasting and has operational application potential.