| 概率匹配订正法在湖北襄阳地区降水预报中的应用 |
| |
| 引用本文： | 袁良,谭江红,闫彩霞,张玉翠.概率匹配订正法在湖北襄阳地区降水预报中的应用[J].新疆气象,2024,18(2):93-99. |
| |
| 作者姓名： | 袁良 谭江红 闫彩霞 张玉翠 |
| |
| 摘 要： | 数值预报产品在天气预报预警中有着重要的作用。2016—2020年汛期ECMWF模式预报降水与湖北襄阳区域站观测降水的对比分析表明：ECMWF对中雨及以上降雨的预报，第1、2天预报偏小，而第3天预报偏大；三个预报时段强降雨中心位置偏差无规律。为了更好地对ECMWF产品进行释用，提高汛期降水预报准确率，从概率匹配角度研究了不同降水量级订正值，并对2021年汛期ECMWF降水预报进行逐日检验。结果显示：概率匹配订正法可有效地改善模式预报性能，对中雨及以上降雨预报均有良好的订正效果，尤其对第1天暴雨预报改进最为明显。228站平均的TS评分提高了6个百分点，由11.1%增加到17.1%，漏报情况改良了13.5个百分点，由85.0%降为71.5%。采用该订正法开展定量降水预报，由于增加了当地降雨概率分布的背景信息，能表现出比原模式更高的参考价值。
|
| 关 键 词： | 降水预报 累积概率 概率匹配 模式订正 |
| 收稿时间： | 2023/2/21 0:00:00 |
| 修稿时间： | 2023/4/25 0:00:00 |
Application of Probability Matching Correction Method to Precipitation Forecast in Xiangyang of Hubei Province |
| |
| yuanliang,tanjianghong,yancaixia and zhangyucui.Application of Probability Matching Correction Method to Precipitation Forecast in Xiangyang of Hubei Province[J].Bimonthly of Xinjiang Meteorology,2024,18(2):93-99. |
| |
| Authors: | yuanliang tanjianghong yancaixia and zhangyucui |
| |
| Abstract: | Numerical forecasting products are essential in weather forecasting and warning. The comparison of precipitation forecast by ECMWF and precipitation observed at Xiangyang Station in Hubei Province during the flood season from 2016 to 2020 shows that the forecast for moderate rain and above rainfall on the first and second days is smaller, while the forecast on the third day is larger. The deviation of the heavy rainfall center in the three forecast periods is irregular. To better utilize ECMWF products and improve the prediction accuracy of precipitation during the flood season, the setting values of different precipitation classes were studied from the perspective of probability matching. The ECMWF precipitation forecast for the 2021 flood season was then tested daily. The results show that the probability matching correction method can effectively improve the model''s prediction performance, and has a good correction effect on the moderate rain and above precipitation forecast, especially on the rainstorm forecast of the first day. The average TS score of 228 stations increases by 6 percentage points from 11.1% to 17.1%, and the missing rate decreases by 13.5 percentage points from 85.0% to 71.5%. The revised method demonstrates higher reference value than the original model because it increases the background information of the local rainfall probability distribution. |
| |
| Keywords: | precipitation forecast cumulative probability probability matching model correction |
|
| 点击此处可从《新疆气象》浏览原始摘要信息 |
| 点击此处可从《新疆气象》下载免费的PDF全文 |
