摘要
该文对2013年昌吉州区域3种主客观温度预报产品进行质量检验,预报时效24 h、48 h、72 h;对温度预报误差≤2℃的百分率（准确率）、平均绝对误差、均方根误差统计检验。对预报员和模式预报产品的质量、技巧评分进行对比分析。结果表明：预报员的24 h预报质量略优于模式的预报质量。48~72 h最高温度预报质量,EC细网格预报最好。分析主客观预报质量的月变化和站点变化,为业务工作中的参考权重提供依据。提出了选取最优订正值订正客观预报系统偏差的方法（预估若干订正值,分别计算每个订正值在一段时间内的平均绝对误差和准确率,选取其中平均绝对误差最小且准确率最高的订正值）,并通过制作相应软件投入业务运用。
In this paper, three kinds of subjective and objective temperature forecast product quality inspection aged 24,48,72 h in Changji of 2013 ;Percentage of temperature 2 ~C or less prediction accuracy ,mean absolute er- ror and root mean square error statistical tests. For forecasters and model forecast the quality of the product, skill score were analyzed. The results show that the forecaster＇s 24 h quality is a little better than that of model prediction quality. Between 48 and 72 hour temperature forecast quality, EC Fine grid forecast is highest. Analysis of the subjective and objective forecast quality and site changes, provide the basis for the reference weight business work. Proposed to select the optimal order when the objective forecast system deviation correction method （estimates several book comes at a time, calculate each order comes in a period of time the average absolute error and accura- cy, select one of the highest average absolute error minimum accuracy order comes at a time） , and by making the corresponding software into business use.
出处
《贵州气象》
2014年第4期1-7,共7页
Journal of Guizhou Meteorology
关键词
温度预报
质量检验
主客观天气预报
业务运用
Frequency
Temperature prediction quality inspection
Subjective and objective weather forecast
Business use
