王佳津, 王春学, 曹萍萍, 陈琳, 冯汉中. GRAPES全球模式预报检验评估——2016年夏季四川地面要素[J]. 高原山地气象研究, 2017, 37(1): 33-40,96. DOI: 10.3969/j.issn.1674-2184·2017.01.006
引用本文: 王佳津, 王春学, 曹萍萍, 陈琳, 冯汉中. GRAPES全球模式预报检验评估——2016年夏季四川地面要素[J]. 高原山地气象研究, 2017, 37(1): 33-40,96. DOI: 10.3969/j.issn.1674-2184·2017.01.006
WANG Jia-jin, WANG Chun-xue, CAO Ping-ping, CHEN Lin, FENG Han-zhong. Assessment of the GRAPES-GFS Forecast Verification about 2m Temperature and 10m Wind in Sichuan Province during Summer in 2016[J]. Plateau and Mountain Meteorology Research, 2017, 37(1): 33-40,96. DOI: 10.3969/j.issn.1674-2184·2017.01.006
Citation: WANG Jia-jin, WANG Chun-xue, CAO Ping-ping, CHEN Lin, FENG Han-zhong. Assessment of the GRAPES-GFS Forecast Verification about 2m Temperature and 10m Wind in Sichuan Province during Summer in 2016[J]. Plateau and Mountain Meteorology Research, 2017, 37(1): 33-40,96. DOI: 10.3969/j.issn.1674-2184·2017.01.006

GRAPES全球模式预报检验评估——2016年夏季四川地面要素

Assessment of the GRAPES-GFS Forecast Verification about 2m Temperature and 10m Wind in Sichuan Province during Summer in 2016

  • 摘要: 本文使用2016年6~8月GRAPES全球模式2m温度和10m风场24、48、72h时效预报场和同期四川省156个国家气象站逐日地面温度和风场资料,选取预报准确率、平均误差、平均绝对误差、均方根误差和Alpha Index5个统计量对2016年夏季四川3个区域(盆地区、过渡区和高原区)2m温度和10m风场进行了较为详细的检验评估。研究结果表明:模式对10m最大风速的预报效果较好,准确率较高,最高可达80.64%。模式对过渡区温度预报效果较差,准确率基本低于10%,但是对盆地区温度的预报有一定可信度。模式对10m最大风速的风向预报效果不如最大风速值。全省各区各要素的AI值都在0.7左右变化,表明模式预报的随机误差大,预报和观测吻合较差。本研究还发现,整体来看模式对盆地各要素预报效果较好,对于地形复杂地区(高原区、过渡区)预报效果较差。此外,模式存在一定的系统误差,2m温度的系统误差盆地区约为-2.3~-1.7℃,过渡区约为-8.3~-6.0℃,高原区约为-7.3~-5.0℃;10m最大风速值的系统误差盆地区约为-1.3~-0.6m/s,过渡区约为-2.3~-1.3m/s,高原区约为-2.7~-1.1m/s。

     

    Abstract: Using the output from the run of GRAPES-GFS for Sichuan province during summer in 2016, its capability to forecast the 2m temperature and 10 m wind in Sichuan (basin area, transition area, plateau area) is evaluated in terms of the forecast accuracy, mean error, mean absolute error and Alpha Index. The results show that the forecast accuracy of 10 m wind velocity is the highest, and the maximum can up to 80. 64%. But the forecast accuracy of 2m temperature in transition area is the lowest, the values are under 10%, but the vales in basin are credible. The 10 m wind speeds of the GRAPES-GFS forecast in Sichuan are better than the 10 m wind direction. The Alpha Index in every area is almost close to 0. 7, which shows that the random error of the GRAPES-GFS is large. The value predicted form the GRAPES-GFS does not tally with the observed value.Overall, this model is effective on predicting the 2m temperature and 10 wind in basin, but less effective in complex terrain area (transition area and plateau area) . Through the verification it is found that there is a certain systematic error in the forecasts, 2m temperature of the system error is about-2. 3 ~-1. 7℃ in basin area, about-8.3 ~-6.0℃ in the transition zone, about-7.3 ~-5.0℃ in plateau area; and 10 m maximum wind speed system error is about-1.3 ~-0.6m/s in basin area, about-2.3~-1.3m/s in the transition zone, about-2.7 ~-1.1m/s in plateau area.

     

/

返回文章
返回