选用VLF/LF(ADTD_2C)三维闪电监测资料和双偏振雷达资料,对2017-2020 年福建省31 个冰雹、32 个短时强降水单体闪电特征进行统计分析.结果表明:福建省多数冰雹云单体降雹前闪电频数超过 50 次/6 min,而强降水单体超过50 次/6 min的较少;冰雹单体正地闪和正云闪比率较高,强降水单体则较低.80%冰雹单体闪电频数峰值较降雹时间提前3～25 min,并在降雹前出现总闪频数跃增,递增率多数大于4 次/min.超过 1/2 的强降水单体,其闪电频数峰值时间较降水峰值提前2～35 min,闪电频数在降水峰值前增大,递增率多数小于4 次/min.两类单体成熟阶段的云闪频数最高,云闪主要集中分布在2～6km高度层.冰雹单体中,融化层以下为冰雹和大雨粒子组成的低层,以上为冰雹和霰组成的高层;强降水单体中,融化层以下为大雨粒子组成的低层,以上为冰晶、过冷水滴组成的高层.闪电频数与强回波中心高度和强回波伸展高度均为正相关,对流发展高度越高,冰相过程越显著,闪电活动越强.

为了深入认识冰雹云中闪电的演变特征及云物理机制,利用VLF/LF三维闪电监测资料,结合S波段双偏振雷达、地面观测等资料,采用统计、对比方法,对福建2017—2020年31个冰雹单体闪电特征进行了分析.结果表明:降雹前闪电频数峰值约有2/3在50次·(6 min)-1以上,80％冰雹云地面降雹出现在闪电峰值后的3～25 min;降雹前总闪电频数出现跃增,70％雹暴单体频数平均递增率达4次·min-1以上,闪电快速跃增提前于降雹前6～40 min;云闪频数在成熟阶段最多,发展阶段最少;冰雹单体三个阶段云闪集中分布在2～6 km高度层;差分反射率因子(Zdr)、相关系数(CC)等参数及粒子相态识别分析表明雹云降雹前融化层以上由冰雹和霰组成,融化层下由干、湿冰雹和雨粒子组成,低层则主要由湿冰雹和中大雨粒子组成;闪电频数、正地闪或正云闪占比率与回波强度、最强回波高度、强回波伸展高度呈正相关.结合了闪电资料与双偏振雷达参量,为识别冰雹云体演变及雷电预警提供参考.

通过自动站、雷达、风廓线雷达等实况资料和欧洲中期天气预报中心(ECMWF)全球模式细网格、0.25°×0.25°经纬度分辨率ERA-Interim再分析、WRF_ARW3.9.1数值模式等模式资料,对2109号南海台风"卢碧"的路径和强度进行诊断和预报偏差分析.结果表明:台风"卢碧"在季风槽中生成,副热带高压、高空槽和双台风等的变化和相互作用是导致其移速缓慢和路径曲折的原因;台风的移动一直没有脱离季风槽控制,总体受西南季风气流引导向东北方向移动;台风强度和结构的变化对台风路径亦有一定的影响.模式对南海台风的路径预报并不因时间的临近而趋于准确,仍存在误差,除了关注模式路径预报,更应分析本身预报形势调整引起的路径变化,也应多分析多参考几家模式.强风法则、变压和南亚高压均可作为台风移向的判别依据.地形敏感性试验表明,山脉地形对弱台风路径有一定的影响.台风强度偏弱,大风风圈半径小,在福建省沿海经历了短暂的减弱(6日5—16时)又增强过程,中部沿海在此时间段出现大风是由不同系统相互作用形成的.热带低压升级的时间偏晚,可考虑在17时前后提前将其升级为热带风暴.

利用2017—2018年葵花卫星(Himawari)TBB亮温资料,计算最低亮温、亮温梯度、红外与水汽亮温差和低亮温区面积及其随时间变化率等特征参量,确定短时暴雨的卫星参数阈值,并融合了雷达参数阈值及过去1h地面加密降水实况资料,采用指标叠加法判定监测区域内某一云团未来2h能否产生区域性短时暴雨天气,并采用交叉相关法外推云团的移动,进而对强降水云团进行预警.对2019年几次暴雨过程预报检验结果是:预警命中率(POD)为80.6％～97.1％,平均为91.0％,临界成功指数(CSI)为77.2％～79.2％,平均为77.9％,所预警的云团未来2 h影响区域出现≥30 mm/h短时暴雨站数占全省短时暴雨站数的76.4％～96.2％,平均为85.2％,整体预警效果较好.

利用南平市275个自动站2013—2018年夏季(7—9月)逐小时降水资料、同期邵武探空资料及ECMWF 1° ×1°再分析资料,统计分析夏季短时强降水的时空分布特征,同时对影响天气系统进行分类,并分析不同天气型的物理量特征.结果表明,南平市夏季短时强降水的频数年际变化大,其日变化呈单峰型,午后到上半夜最易发生短时强降水,峰值出现在17时,易发地段主要集中在东南部及西北部;影响系统主要分为副高型、冷槽型、暖切型及热带低值系统型,其中副高型出现次数最多,同等条件下,热带低值系统型天气最易发生短时强降水,副高型天气最不易发生短时强降水.

利用VLF/LF三维闪电监测定位资料,结合雷达、地面观测等资料,采用统计分析、对比分析法对福建2017—2020年33个冰雹云的闪电特征进行分析.结果表明:降雹前闪电频数峰值约有66.67％在50次/6 min以上,且降雹前总闪电频数出现跃增;70％雹暴单体平均递增率达4次/min以上,闪电快速跃增提前于降雹时间大多在10 min以上;成熟阶段云闪/总闪比值是3个阶段中最大,成熟阶段云闪是3阶段中最多的,初生发展阶段最小;80％冰雹云地面降雹落后于闪电峰值出现后的3～25 min,滞后于峰值时间最长超过30 min.这些结果可以为冰雹云的识别预警及人工防雹作业指挥提供参考.

2014年3月26-29日,福建省出现连续强对流天气过程,分析表明,大尺度环流背景给此次强对流天气连续发生发展提供了有利的条件.物理量诊断分析,同时参考历史统计的各参数的阈值,对判断强对流天气发生发展、分类和落区有帮助:福建上空大气层结不稳定,中低空存在中等到强的垂直风切变;前期比湿9～10g/m3左右,后期增大,全省大部在12g/m3左右,前期为干对流型,后期为混合对流.强对流天气临近预报,要注重卫星云图上游地区云系的发生发展状况以及与已经发生的天气实况的对应关系,加强多源资料的监测应用,从而提高强对流天气临近预报的提前量.

The temporal and spatial distributions of convection systems(CSs) that influence the Yangtze-Huaihe River Basin, and the param-eter characteristics of mesoscale convective systems from 1998 to 2007 were statistically analyzed by using Deep Convection Tracking Prod-ucts provided by the International Satellite Cloud Climate Program(ISCCP).The results showed that these CSs are mainly concentrated in the spring and summer, mostly generated in the Yangtze-Huaihe River Basin and central and western regions of China. These CSs exhibit a zonal distribution characteristic with the center at the Yangtze-Huaih River Basin;the closer the area to the basin, the denser the CS distribution is. The number of CSs is much larger than other large scale convections. According to their origins, the CSs can be divided into 5 categories. Af-fected by climate conditions and topography, these CSs are significantly different from other each. Overall, the greater the horizontal scale of CSs is, the greater the life cycle, the number of convective clusters (CCs) and the cloud temperature gradient are. Among them, the average values of the horizontal scale, the life cycle and the number of CCs in MID region are the smallest. In the SE region, CSs are mainly with medi-um and long life history, where the average values of the horizontal scale, maximum convective ratio and the cloud temperature gradient of CSs are the largest. During the rainy season in the Yangtze-Huaihe River Basin, convective activity is frequent. The horizontal scale and life histo-ry of CSs are long, and there is a large number of CCs within CSs.

利用地面逐时降水资料、常规天气资料、雷达、卫星以及台风年鉴资料,对39个登陆福建不同地段的台风短时强降水空间、时间分布及强降水强度、落区、过程雨量分布及其与登陆时台风强度关系等进行统计分析,揭示登陆福建台风短时强降水规律,结果表明:(1)登陆中部台风强降水站数虽多,但单站强降水持续时间最短;登陆北部台风强降水站数虽少,但单站强降水持续时间最长.(2)强降水主要时段登陆北部、中部类是登陆前5小时至登陆后5小时,登陆南部及南海北上类是登陆时至登陆后15小时.此外,南海北上类及登陆北部类强降水落区多数在内核区;登陆中部类强降水落区在台风中心附近的频率是几类中最低的,大部分强降水落区在北侧的螺旋雨带上.

In August,2014,No.2 tropical depression (TD2)moves northward after landing in Xiapu Coun-ty,Fujian Province.It results in an extremely heavy rainstorm in northwestern Fujian Province and the related flood and geological disasters.These disasters affect nearly 100,000 people with 4 people dead,and cause an economic loss of 250 million RMB.Great discrepancies in terms of the intensity and location of this rainstorm are found a-mong different numerical prediction models (NWPs).To analyze this extremely heavy rainstorm process,various data including automatic meteorological station,radar products,and routine weather map are used.It is shown that this rainstorm is attributed to the interactions between northward-movement TD2 arising from the South China Sea and cold air from high latitudes.The rainstorm originates from this type of weather background:eastward migration of high-level trough,wind shear in middle-low levels,southward movements of cold front,and northward move-ments of TD2 in the South China Sea.After landing in Fujian,the maintenance of TD2 intensity is due to the incor-porations of the moist baroclinic frontal zone related to the cold air,positive vorticity area near the wind shear in middle-low levels,and high divergence area before high-level trough and to the right of the southwesterly jet. Northward-movement TD2 provides a great amount of water vapor with high energy and high temperature.Through interacting with the cold air,unstable condition and ascending motion intensify,which further triggers the convec-tion.Since the cloud clusters of the TD2 and the westerly trough merge in Northwestern Fujian Province,a more powerful cloud cluster is formed.The location of the extremely heavy rainstorm is in good accordance with the area of moist baroclinic front,convective unstable zone,overlapping area of positive moist potential vorticity in high lev-el and negative moist potential vorticity in low level,and ascending motion zone.

The severe convective predictive equation is set up with the method of stepwise discrimination analysis by using the TBB data from 2005 to 2008,to decide if a cloud cluster in the monitoring area will produce regional severe weather with the indicator superposition method.Based on the statistics of radar parameters during hail and gust events with thunderstorms for many years in Fujian,the radar parameter thresholds for severe convective weather,such as big hail,hail,and gust with thunderstorm,are decided.According to the height difference between the height of the 0 ℃ level and the storm top and that between the height of the-20 ℃ level and the storm top,the storms captured by CINRAD are chosen one by one,so to identify the area that will be influenced in the next one hour by storms that may bring about big hail,hail or gust weather.

Based on the radar mosaic 3D data with clutter elimination algorithm and an extensive quality control program developed by the State Key Laboratory of Severe Weather,Chinese Academy of Meteorological Sciences,a methodology is presented for the automated identification,tracking and nowcasting of storm.The convective storms of each plane are detected using the region-growing method based on constant altitude surfaces(CAPPI) between 0.5 km and 16 km of radar mosaic,which are associated in the vertical direction in order to get whole storm.The characteristic parameters of whole storm such as storm center,top,base,VIL,volume,moving vector etc.are then computed.The detected storms are tracked in successive images based on combinatorial optimization method,with some geometric logic to deal with merges and splits.Then forecasts are made for the several parameters such as storm center,projected-area ellipse and VIL.The performance of the detection and nowcasting are evaluated by applied in several convective cases.It shows that the identification method could achieve automated detect storm,the volume change is the crisis of main features of different stages of development from the storm characteristic parameters change,while the storm,the average reflectivity factor and maximum reflectivity factor variation with time is not obvious.It is considered that the tracking is relatively reasonable even in the case of merge and split by analyzing the observational radar data.The forecast position of storm is better than the size and VIL,mainly due to the storm development and change constantly,as well as storm merge,split occurred frequently.It is very difficult to forecast the future development trend only based on its history.Overall,the prediction error increases with the forecast period increasing.

The forecasting equations of severe convective weather were built using the physical variables obtained from Model MM5 and radiosonde data during 2003 to 2007,by the ways of making the statistic space distribution with location of severe convection and collocating parameters.By considering the seasonal variation characteristics of the factors,potential forecasts of severe convection and location are made in the future 0-6 h and 0-12 h.The results show that the skill of potential forecast is higher during March to June than July to September,and regional than non-regional.The method is of good instructive sense to forecast severe convective weather of hail,disastrous gust and so on,although some False Alarm Rate events are produced.

By using radar data of Longyan CINRAD/SA from 2003 to 2007,the mesocyclone products were statistically analyzed.The characteristics of mesocyclones determined by person and the persistent three volume scannings and corresponding storm are mainly analyzed.The analysis shows that the mesocyclone which has more than three volumes scanings is well related with strong weather phenomena such as hail, thunderstorm,short-time strong rainfall and so on.Based on the analysis of the typical strong weather processes,we can conclude the variation rules of the mesocyclone height and the largest shear's height, and the correspondence relationship among the shear,the strong or weak mesocyclone and different types of strong weather.Therefore,the above conclusins will provide references for forecasters to predict short-time strong rainfall,hail,thunderstorm timely and accurately.

Based on the routine meteorologic data and air sounding data from 2003 to 2007 in Northwest Fujian,the characteristics of the leading weather systems influencing on local precipitation were analyzed.The result shows that,low-vortex shear is the chief precipitation weather system in Northwest Fujian,and then are continental-high base and warm sector convergence.There are visible differences in the precipitation distribution of each grade for the 10 kinds of weather systems.The opportunity of storm rain is largest when the weather is controlled by low-pressure inverted trough.Northwest Fujian is affected most by the southern part of low-vortex shear,of which the southern,middle and northern parts are the positions relatively suitable for local artificial precipitation operation.The occurrence rate of storm rain is great when the weather is controlled by low-vortex shear in spring,so it should be circumspect when carrying out artificial precipitation operation.The height of 0℃ layer is below 4200 meter on average,and the monthly distribution changes as unimodal shape,for which January is the lowest while July the highest.

Based on Doppler radar scan data and automatic weather station data from 2002 and 2003 and the statistical analysis method,the ZI relation formulas for different regions and different sorts of rainfall in the northern and central Fujian Province are obtained.Precipitation processes in 2005 and 2006 are calibrated by means of the relation formulas.The variational method and optimum Kalman Filter method(in combination with Kalman Filter and optimum interpolation) are used to calibrate the estimated rainfall with real-time data,and the calibrated rainfall is tested with the raingauge data from the Jiuquxi Valley of Wuyishan Mountain.Results for the methods are compared and analyzed.The analysis shows that the smallest error results are from the optimum Kalman Filter method and optimization method from 0 to 100 km,and the greatest error is resulted from the relation of Z=300I1.4.

A squall line attacks Jiangxi,Zhejiang and Fujian provinces on April,12 2003.The severe convective system leads to heavy damage,producing wind gust as large as 32 m/s,30 mm diameter large hails in the affected area.This process is observed by the CINRAD in Jianyang City.The whole process of this convective system,including the initial stage,the developing stage,the mature stage and the dissipating stage,is analyzed.Many convective structures are observed in the process.At the location of the strong updraft,base reflectivity observed by midlevel elevation is stronger than the low-level and the high-level.This structure is called weak echo area.Corresponding to weak echo,there is a positive velocity area within a large negative velocity area in the mid-level.This structure is judged as mesoscale vortex sometimes by WSR-98D PUP.In fact,this phenomenon means the convergence in the target area,which is known as MARC(mid-level radial convergence).Strong line echoes are observed in squall line by radar.It is well known that bow echoes that develop within a squall line are referred to as line echo wave patterns.In order to analyze the interior structure of the convective system,mesoscale numerical model(MM5V3) is used to simulate this process.In the control numerical simulation,reanalysis NCEP data(1°×1°) is used as the initial conditions.Radar observation data and simulation result are used to analyze the structure and the evolvement.Numerical simulation result confirms that the convective system takes place at the meeting of the northeaster and the southwester.There are many convective vortices and super cells within the 400 km system.Bow echoes appear near the vortices.There are obvious cyclones at the forehead of the line echo and divergence at the tail.Weisman indicates that in a mature bow echo cyclonic and anticyclonic vortices develop north and south respectively,in the channel of rear-to-front flow.These results resemble Weisman's study about squall lines in American.In the period from 0758UTC to 0840UTC,there are respective bow echoes in the north and west,seen from the reflectivity PPI.These results confirm that there are segments of mesoscale vortices and bow echoes in squall lines.Jianyang city is to the east of Wuyi Mountain,and there is a bell-mouthed mountain to the west of Jianyang city,so the influence of the topography should be considered.In order to do that,a sensitive numerical simulation is designed.In the sensitive numerical simulation,the altitude of the mountains near Jianyang is factitiously set equal to the ground.The results of the control and sensitive numerical simulation are compared.It shows that the uplifted velocity in 500 hPa level of control simulation is greater than the velocity of sensitive simulation,and the uplifted area accords with the strong reflectivity echo.These results refer that the windward mountain and bell-mouthed mountain can trigger new convective cells or enhance existing ones.

Based on the visual fortran and visual basic programming language and the related algorithm, precipitation software system of the Doppler radar quantitative measurement to Jiuquxi,Mt. Wuyi is developed with the 28 processes echo data of Jianyang CINRAD radar from 2002 to 2004, and the precipitation data of the 27 surface meteorological stations in the mid-north Fujian, and the 10 automatic rainfall stations data of Jiuquxi in the Mt. Wuyi. The 9 groups optimal A and b value of three kind of different precipitation types in three different regions are determined. The effect of the three kinds of precipitation products in precipitation quantitative measurement is more obvious than that of the PUP of Jianyang CINRAD radar.

Based on the Doppler radar data and the data from twenty-seven automatic weather stations from May to June in 2002 and 2003 in the northern and central Fujian,the statistical analysis of optimization is made to obtain the Z-I relation of different regions.The rainfall amounts in precipitation processes in 2005 are estimated.Using variational method and optimum Kalman filter method (combine Kalman Filter with optimum interpolation), radar estimations are tested with rainfall data of Wuyishan Jiuquxi Valley in the northern Fujian.The accuracies for the methods are compared and analyzed.The results show that mean relative errors are less than 25% with variational method and optimum Kalman Filter method,it is also less than 25% from 0 to 100km with optimization method,however,it increase remarkably from 100 to 200km.Mean relative error is the greatest with the relation of Z=300I 1.4.