| 李淑娟,毛炜峄,于晓晶,陈颖,贾孜拉.拜山.基于DERF2.0的乌鲁木齐春季降温过程的延伸期预报检验[J].沙漠与绿洲气象,2018,12(3):40~48 |
| 基于DERF2.0的乌鲁木齐春季降温过程的延伸期预报检验 |
| DERF2.0 Extended-range Assessment for Urumqi Spring Cooling Process |
| 投稿时间：2018-01-02 修订日期：2018-01-25 |
| DOI：10.12057/j.issn.1002-0799.2018.03.006 |
| 中文关键词: DERF2.0 预报检验 偏差 预报时效 强降温过程 |
| 英文关键词:DERF2.0 Evaluate the forecast effect Bias Forecast time Strong cooling process |
| 基金项目:中央级公益性科研院所基本科研业务费专项(IDM2016003). |
|
| 摘要点击次数: 974 |
| 全文下载次数: 449 |
| 中文摘要: |
| 基于DERF2.0数据,应用均一化标准差、均方根误差等方法,以2013年乌鲁木齐春季逐日温度及24h变温检验为背景,初步评估了该模式对延伸期春季强降温过程的预报能力。结果表明：(1)逐日气温预报整体偏差随预报时效推进而增大,延伸期预报偏差明显大于中期。(2)旬平均温度的中期预报偏差普遍在-2～-8℃,延伸期的预报偏差最小在0℃左右,最大为-15.5℃。(3)日平均气温以及最高、最低气温的逐日偏差均以冷偏差为主,偏差范围为5～-15℃,延伸期预报偏差范围为-5～-15℃。模式对升温阶段的预报冷偏差随升温加剧而增大,对降温阶段的预报偏差随降温加剧而减小。24h变温偏差主要在5℃范围内变化,延伸期的24h变温偏差比中期预报偏大可达|8|℃以上。(4)DERF2.0模式对中短期温度预报有一定水平,延伸期预报能力下降,可参考价值较弱。(5)对强降温过程的结束日的温度预报偏差小,而对过程初始日的温度预报冷偏差大,造成对降温过程的预报暖偏差大,强降温过程普遍漏报。 |
| 英文摘要: |
| Based on DERF2.0 data, use the method of Normalized Standard Deviation (NSTD) , RMSE and so on, with the background of Spring daily temperature and 24h temperature-change evaluation in 2013, A preliminary assessment of DERF2.0 extended-range prediction ability for spring strong cooling process is made. The results show that: (1) The overall daily temperature bias is increased with impel of forecast time, extended-range forecast bias is more than medium-range obviously. (2) The bias of ten-day average temperature generally in -2~-8℃, extended-range forecast minimum bias is about 0℃, the maximum bias is -15.5℃. (3) Daily bias of 2mT、Tmax and Tmin is mainly cold, concentrated in 5 ~ 15℃, extended-range forecast bias is mainly change between -5 ~ 15℃. Cold bias of warming stage is increased with temperature rise, but the bias of cooling stage is decreased with temperature drop.24h temperature-change bias is mainly changes in 5℃, the extended-range forecast bias is |8|℃ larger than medium-range. (4)The DERF2.0 model has certain prediction ability to the mid-shot-term temperature, but extended-range forecast ability is too weak to reference.(5)The ending-T forecast bias of strong cooling process is small, but the cold initial-T bias is large, it made a large process warm bias and missing report. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
|
|