【作者】 束宇；

【导师】 伍荣生； 潘益农；

【作者基本信息】 南京大学 ， 气象学， 2011， 博士

【摘要】 中尺度对流系统(Mesoscale Convective System,简称MCS)是我国夏季造成暴雨和洪涝灾害的主要天气系统之一,静止卫星红外云图是监测MCS的重要手段。因而基于静止卫星红外云图对MCS进行普查是MCS研究的一个重要方向。静止卫星原始数据经过定位、插值、定标三个步骤可以转化为格点上的物理量数据,从而可以用于科学计算中。以往对MCS的普查工作中,由于人工识别MCS的普查方法的限制,对MCS的普查工作往往局限在一个较小的时间跨度和空间范围内,而且带有一定的主观性,不利于对MCS气候学特征的研究。本文中,作者开发了一种基于静止卫星红外云图的MCS的计算机自动识别方法。自动识别方法主要包括三个步骤：MCS轮廓的生成、对MCS的追踪、对MCS的判别。对1999年中国及邻近地区MCS的抽样普查结果表明,自动识别方法在查找MαCS时的错识别率大约为19.1%,误识别率为1.4%,漏识别率为零；查找MβCS的错识别率为10.1%,误识别率和漏识别率均为零。基于新的MCS自动识别方法,作者开发了一款卫星云图资料处理软件：“静止卫星云图处理和强对流侦测系统”。采用基于静止卫星红外云图的MCS的计算机自动识别方法普查了1995～2008年夏半年亚洲和西太平洋地区的四类不同尺度和形状的MCS,共得到47468个满足要求的MCS。然后,以这些普查到的MCS为样本,研究了亚洲和西太平洋地区MCS的形状、尺度、持续时间、移动速度、地理分布、月变化、日变化生命史等的气候学统计特征。结果显示：MCS近线形的数目是近圆形的2.5倍,近线形MCS的尺度更大,近圆形MCS的持续时间更长。500hPa引导气流对MCS的移动起重要作用,MCS成熟以后移动速度会加快。四类MCS的地理分布特征、月变化特征较为相似。MCS基本呈纬向带状分布,从南向北依次有三条活动强度递减的MCS分布带。各地区MCS的月变化与其所处的大尺度环境的季节调整相关联。MCS按日变化特征可以分为四类：低纬全天候MCS、高纬全天候MCS、夜发性MCS和午后形成的MCS。在中国地区,高原和丘陵地区的MCS是单峰型生命史,下午形成,傍晚成熟,入夜后减弱消散；平原地区MCS的生命史是多峰型特征,下午和夜间都有MCS形成；洋面上午夜时是MCS的形成高峰；两广沿海受海陆风环流影响,MCS集中在下午三四点左右形成；四川盆地受山谷风环流影响,MCS夜发性特征显著。利用2008年华南暴雨实验期间的LAPS分析资料、MT1R红外云图资料、多普勒雷达资料对2008年6月12日的一次暴雨MCS过程进行了诊断分析,研究了其发生的环境背景、红外云图上MCS演变过程及对流系统中风暴的发生发展和演变。结果显示：该次暴雨MCS形成于有利的对流发展环境中,低层的西南涡在MCS的形成和演变中起到了重要作用；红外云图显示对流系统是一次MβECS过程；雷达图象显示对流系统是一次飚线强风暴过程,入流区低层反射率因子梯度很大,存在弱回波区,中高层的回波悬垂结构明显,径向速度图显示入流区中层存在辐合区；LAPS资料分析表明强风暴形成于中等的CAPE值和较强的垂直风切变环境中。使用WRF3.2对2009年8月16日～18日发生的一次暴雨灾害过程进行了数值模拟,该次模拟较好的模拟出了该次暴雨过程中的MCS活动和MCV活动。结果显示：MCV形成于800hPa以下,以850hPa涡旋特征最为明显；涡旋形成于对流性降水区中,涡旋形成时涡旋区域内有正负变涡耦极子,涡旋向正变涡的方向移动；在对流层低层,辐合辐散项和倾斜项是正涡度的来源,而对流项和平流项则抑制局地涡度的增大。大气对对流加热的平衡响应,是该MCV的形成机制之一,该MCV形成的另一个机制为对流内上升运动造成的水平涡管的倾斜。对MCV形成时的Rossby形变半径的计算表明,扰动尺度大于Rossby形变半径,即扰动为“动力大尺度”的,MCV将会持续较长的时间。更多还原

【Abstract】 Mesoscale Convective Systems (MCSs) cause heavy rain and other severe weather events during the warm season. Geostationary satellite infrared imagery with high spatial and temporal resolution can provide much available information for MCS surveillance. Hence, MCS census by means of satellite infrared imagery is one of the important aspects of MCS study. After orientation, interpolation and calibration, the raw satellite data can be translated into the grid data, as could be used in scientific computing.In previous studies on MCS census, because of the limitations of MCS identifying method, the census were confined to either a relatively small space and time frame or a particular type of MCS, which was bad for the research on MCS climatic characteristics. In this article, a new method of identifying MCS from infrared imagery by computer is developed, which contains three main steps:forming MCS profile, tracking the MCS, and judging the system according to MCS definitions. Results of the MCS census over China and its vicinity of the year1999show that, the wrong rate of the new method in identifying MaCS was lower than20%, which was about10%while identifying M(3CS. Based on the new method, the author developed a software, Satellite image and MCS plot.The automatic MCS identification method was used to capture four categories of MCS with different size and shape from satellite infrared numerical data. Forty-seven thousand four hundred and sixty-eight MCSs were found over Asia and western Pacific region during the warm season (May to October) from1995to2008. From this database, MCS characteristics such as shape, size, duration, velocity, geographical distribution, intermonthly variation, diurnal variation, and lifecycle were studied. The results indicated that linear MCSs were2.5times the number of circular MCSs. The former was of a larger size while the latter was of a longer duration. The500hPa steering flow played an important role in MCS movement. MCSs tended to move faster after they reached the maximum extent. Four categories of MCS had similar characteristics of geographical distribution, intermonthly variation. Basically, MCSs were in zonal distribution, with three zonations weakening from south to north. The intermonthly variation of MCS in one place was related to the seasonal adjustment of its large-scale circulation. MCS could be divided into four categories according to their diurnal variation characteristics:low latitude all-day MCS, high latitude all-day MCS, nocturnal MCS, and afternoon MCS. As far as the MCSs over China were concerned, they had different lifecycle characteristics over different areas. MCSs over plateau and hill areas, which had only one peak in their lifecycle curves, tended to form in the afternoon, mature at nightfall, and dissipate at night. On the other hand, MCSs over plain areas, which had several peaks in their lifecycle curves, might form either in the afternoon or at night, whereas MCSs over the oceans tended to form at midnight. Affected by sea-land breeze circulation, MCSs over coastal areas of Guangdong and Guangxi Province always came into being at about three or four pm local time, while MCSs over the Sichuan Basin which were affected by mountain-valley breeze circulation generally initiated nocturnally.By means of the LAPS data, MT1R satellite infrared imagery, and Doppler radar data, the process of heavy rain on June12,2008, was analyzed. The background environment, MCS evolution, as well as the formation and evolution of the storm in the MCS were researched in details. Results showed that:the MCS was formed in a favorable environment, and the southwest vortex in low levels played an important role in MCS evolution. The MCS was a MβECS from the satellite infrared imagery and a strong storm of squall line from radar imagery. The reflectivity gradient in low levels of inflow area was large, with a weak-echo region. The overhanging of echo in mid levels was quite clear. The radial velocity image showed that a convergence area existed in the mid levels of inflow area. The strong storm was formed in the environment with moderate CAPE and strong vertical wind shear.The rainstorm during August16th～18th,2009, was simulated by WRF3.2. The simulation succeeded in simulating the MCS and MCV activities during the rainstorm process. Results showed that:MCV formed beneath800hPa, and the rotation was most evident on850hPa. MCV formed in the convective precipitation area. When MCS formed, a vorticity variation couplet existed in the vortex area, and MCV moved along the direction of positive vorticity variation. In low levels of troposphere, the stretching term and tilting term were main source of positive vorticity when MCV formed, and the horizontal advection term and convection term restrained the growth of local vorticity. About this MCV case, one of the MCV formation mechanism was the atmospheric balance response to convective heating, and the other was the tilting of the horizontal vortex tube. After calculating the Rossby deformation radius at the time of MCV formation, the horizontal scale of the circulation was found to be larger than the Rossby deformation radius. It means that, the perturbation was "dynamically large", and the MCV was expected to be long lasting.更多还原