|
|||||
|
|
| “雨燕”中风暴算法在北京奥运天气预报示范项目中的应用及改进 | |
| 引用本文： | 胡胜,汪瑛,陈荣,何如意.“雨燕”中风暴算法在北京奥运天气预报示范项目中的应用及改进[J].高原气象,2009,28(6):1434-1442. |
| 作者姓名： | 胡胜 汪瑛 陈荣 何如意 |
| 作者单位： | 1. 广州中心气象台,广东,广州,510080;中国气象局,国家气象中心,北京,100081 2. 广州中心气象台,广东,广州,510080 3. 广州市番禺区气象局,广东,广州,511400 |
| 基金项目： | 中国气象局新技术推广项目，广东省气象局科技项目 |
| 摘 要： | 介绍了临近预报系统“雨燕”中的风暴系列算法， 包括风暴识别、 风暴追踪、 基于TREC技术的风暴位置预报, 以及预报位置实时评分算法等。利用2008年北京奥运期间出现的多个强对流天气， 统计出该风暴产品在30 min和60 min预报时效的绝对距离误差分别约为13 km和23 km。分析了北京奥运天气预报示范项目期间该风暴产品误差较大的原因， 主要集中于TREC技术本身及其适用范围， 以及风暴预报方案中处理细节的不足， 具体为TREC技术不适用于孤立的回波单体， 雷达探测边界对TREC技术的影响， TREC矢量有时呈现出一定的空间不连续性， 以及对孤立少动单体的不当处理等。针对上述原因， 提出了一系列的改进方案， 包括对用于风暴位置预报的TREC矢量增加一致性检验， 利用风暴的历史轨迹校正不恰当的TREC矢量； 对TREC技术中象素阵列大小进行统计分析， 选择最适合北京地区的阵列大小； 风暴预报位置超出回波范围时新的处理技巧等。利用北京奥运期间的强对流资料， 对改进后的风暴算法进行评估。结果表明， 一方面， 改进后的算法能较好地控制风暴预报位置的绝对误差， 在30 min和60 min预报时效， 绝对误差分别减小了25%和26%。另一方面， 由于预报位置精度的提高， 能够提升相邻时刻风暴匹配的效率， 使得与以前算法相比有更多的风暴样本参与了各个预报时效的评分。 |
| 关 键 词： | 风暴算法 评分 平均绝对误差 |
Applications and Improvements of Storm Series Algorithms in SWIFT during B08FDP |
|
| HU Sheng,WANG Ying,CHEN Rong,HE Ru-yi.Applications and Improvements of Storm Series Algorithms in SWIFT during B08FDP[J].Plateau Meteorology,2009,28(6):1434-1442. | |
| Authors: | HU Sheng WANG Ying CHEN Rong HE Ru-yi |
| Abstract: | Firstly,the storm series algorithms in the SWIFT,including storm cell identification,storm tracking,storm forecast and forecast position verification,are introduced. During the Beijing 2008 Forecast Demonstration Project (B08FDP),these algorithms have been applied. It is found that the mean absolute errors for 30 and 60 min storm forecasts are about 13 km and 23 km,respectively. With the increase of forecast length,the mean absolute errors of the storm product become larger. Secondly,the reasons to cause errors are discussed. The vectors calculated by the TREC technique are not suitable to forecast isolated storms. The quality of TREC vector is limited on radar detecting border. Discontinuity of TREC vector occurs sometimes. And some forecast techniques should be adjusted in details. Thirdly,the improvements of the storm forecast scheme are newly designed,including consistency check for TREC vector by using the storm history vector,statistical analysis on array pixels for the TREC technique to choose the best array length,new processing method for the border or outside of radar detection. Finally,the storm products are recalculated by using the improved storm forecast scheme. The improved scheme performs better than before. The mean absolute errors for 30 min and 60 min are reduced 25% and 26%,respectively. And the storm tracking algorithm is also benefited from the improvements of storm forecast scheme. |
| Keywords: | Storm algorithms Verification Mean absolute error |
| 本文献已被 万方数据 等数据库收录！ | |
| 点击此处可从《高原气象》浏览原始摘要信息 | |
| 点击此处可从《高原气象》下载免费的PDF全文 | |