贵州最高气温预报方法研究

最高气温预报一直以来是贵州最为棘手的问题，近年来在国家气象局预报质量通报中，成绩较为靠后。为改变这一现象，笔者基于各预报中心常规气温预报资料及地面观测资料，在贵州境内展开最高气温的多模式集合预报研究。结果表明，多模式集合预报技术有效地改进了预报的准确率，在对2013年1月1日—2014年4月30日120 h的逐24 h预报中，各预报中心的多模式集合预报结果明显降低了预报的均方根误差，效果远优于最好的单个预报中心(ECMWF)和多模式的集合平均，不仅很好地改善了贵州最高气温的预报效果，还给当地预报及决策气象服务提供更有效的参考。

Forecast of Maximum Temperature in Guizhou

It is a tough issue to forecast the maximum temperature forecast in Guizhou, and the forecast quality was not satisfied in recent years. Based on the conventional temperature forecast data and ground observation data from forecasting centers, the author carried out the multi- model ensemble forecast for the maximum temperature. The results showed that: the multi-model ensemble forecast could effectively improve forecast accuracy, and obviously reduce the forecast RMSE of hourly forecast in 120 hours from 1 January, 2013 to 30 April, 2014. This technique had a considerable advantage over the best single model forecast (ECMWF) andthe ensemble mean (EMN). The study could improve the maximum temperature forecast in Guizhou and provide more effective guidance for local meteorological forecast and decision-making.