一种新的梅雨锋上中尺度涡旋识别方法

中尺度涡旋的发生、发展对梅雨锋暴雨常具有直接作用,客观准确地识别中尺度涡旋有助于提高暴雨预报的准确性。本研究提出一种从格点风场中自动识别中尺度涡旋中心的客观方法。利用美国国家环境预报中心(NCEP)提供的全球模式分析资料,选取2013—2014年梅雨期间两次暴雨个例,考察新方法识别中尺度涡旋的能力,并与现有的两种识别方法(分别基于相对涡度场与基于高度场)进行比较分析。结果表明,由于较小尺度的系统不遵守地转风规则,梅雨锋上许多涡旋的风场环流中心、涡度中心与低压中心位置不重合,影响通过涡度识别或气压识别方法的准确性。新方法从风场出发,可准确识别出大多数涡旋中心,误判率低,定位精度高于无人工辅助下的另外两种方法。接着利用新方法分析了两次暴雨个例中不同中尺度涡旋的垂直结构与时间演变。分析表明,新方法无需人工辅助,无特定层高和时间限制,可在短时间内识别出区域内所有中尺度涡旋的位置、三维结构与时间演变,可用于梅雨期间静止锋上中尺度涡旋的识别和路径的追踪,有助于预报员实时分析与预报暴雨。

A New Detection Algorithm of Mesoscale Vortex During Meiyu Period

The occurrence and development of mesoscale vortices directly influence heavy rainfalls during Meiyu period. Identifying mesoscale vortices accurately helps to improve the prediction of heavy precipitation. An objective method is proposed to automatically detect the centers of mesoscale vortices from gridded wind field in this study. Using global analysis from National Centers for Environmental Prediction (NCEP), USA, two cases during the 2013-2014 Meiyu periods were selected to compare the existing two algorithms which are based on vorticity or height distribution and the new one. The results indicate that since small scale systems do not obey geostrophic wind relationship, the centers of wind, vortex and low pressure scarcely overlap Meiyu fronts, thus decreasing the accuracy of detection on the basis of vorticity or pressure. However, the new method is based on winds and can precisely identify most vortex centers with a higher precision than the other two algorithms. Then the new method is used to analyze the vertical structure and time evolution of different vortices in the two cases. The analysis shows that the new method does not need any manual intervention, specific level or time restriction. It can identify the position, 3 D structure and time evolution of any mesoscale vortex falling in the domain in a very short time. Also, it can be used to detect mesoscale vortices on stationary fronts during Meiyu periods, very helpful for forecasters to do real time analysis and forecast severe precipitation.