新疆冰雹分区预报方法研究

利用1981-2011年的气象观测数据,普查分析了位于新疆北疆沿天山一带的奎屯河-玛纳斯河流域和位于南疆中天山南麓的阿克苏河流域两大防雹区长序列降雹个例的环流形势背景和天气条件,把两地的主要降雹环流型归纳为中亚低涡(槽)型、巴尔喀什湖低涡(槽)型、西西伯利亚低涡(槽)型和锋区短波槽型等4种类型,并根据两地各环流型降雹的三度空间配置及环境条件,研制出了6种降雹概念模型.同时,筛选出8个有指示意义、稳定可靠的对流因子作为冰雹预报因子,建立了两地各环流型多指标叠套法的冰雹分区、定量(强)预报方法.该方法实际应用效果显著,可以推广到新疆其他重点降雹区域应用.     

新一代天气雷达常用产品在我国人工影响天气工作中的应用

随着新一代天气雷达的组网建设,常用雷达产品逐渐在人工影响天气（以下简称＂人影＂）工作中发挥了重要作用。如利用反射率因子产品分析降水系统的移向、移速、强度以及降水性质等特征为人影作业指挥提供参考;利用径向速度产品选择人影作业的最佳时机和部位;利用垂直积分液态水含量产品识别冰雹云并指导防雹作业;综合应用多种雷达产品建立人影...     

雷达识别冰雹方法的进展

回顾了用气象雷达识别冰雹云方法的进展情况，介绍了双波长雷达、园偏振雷达，特别是双线偏振雷达在研究云体中降水粒子相态方面的原理及部分观测结果。     

1970-2013年南疆渭干河-库车河流域绿洲冰雹时空变化特征和棉花受灾分析

渭干河-库车河绿洲是南疆地区主要棉花生产基地之一, 随着气候变化和棉花种植面积的日益扩大, 雹灾对棉花生产的危害日趋严重. 利用该地区4个气象站44 a逐一地面降雹过程资料和33 a灾情信息, 对冰雹时空变化特征、活动规律、强度、受灾面积和影响因素等进行研究. 结果表明: 降雹空间分布为北部靠山地带多于南部平原地带, 受灾面积的空间分布则相反; 降雹以周期短、强季节性特征为主, 各县中等以上降雹天气占80%以上, 降雹趋于高发和强发的态势. 降雹频次与气温和降水呈较好的相关性, 冰雹多发年气温偏低、降水偏多, 冰雹少发年气温偏高、降水偏少. 位于南部靠近沙漠的沙雅县受灾最为严重, 雹灾与降雹强度、地域、季节和作物面积及类型等有关, 该地区棉花受灾机率大.     

新疆喀什西部新一代天气雷达产品的冰雹预警指标研究

2010年以来喀什地区冰雹频繁,严重危害当地农业生产. 通过整理分析2009-2012年喀什新一代天气雷达监测资料,结合我国中东部地区基于雷达监测的冰雹预警指标,经过对喀什西部的18次冰雹天气过程中新一代天气雷达监测预警指标的分析,归纳出了喀什西部新一代雷达产品的冰雹预警指标：（1）组合反射率≥50 dBz;（2）回波顶高≥9 km;（3）云体垂直液态水含量≥20 kg·m^-2. 根据该预警指标对喀什西部2011-2012年冰雹过程进行回报预警,准确率达到84.6%. 以有详细降雹时间记录的14次冰雹过程为例,分析喀什西部基于雷达监测产品的冰雹预警时效,14次个例中冰雹预警时效平均为27 min. 选取2012年5月13日、5月23日和2013年6月18日3次强冰雹过程个例,分析冰雹出现前后喀什新一代天气雷达的连续监测资料表明：3次过程中降雹开始时间对应雷达监测值达到最大（高）值时刻;组合反射率与回波顶高首先达到预警指标,垂直液态含水量达到预警指标的时间滞后11~19 min;3项雷达产品均达到冰雹预警指标的时间比实际发布冰雹预警时间早7~17 min,喀什西部冰雹预警时效潜力可以增加到30 min以上. 在喀什西部,应用新一代天气雷达监测产品开展冰雹实时预警业务是可行的,对提高当地冰雹灾害防御能力提供了有力的技术支持.     

地质雷达在地下埋藏物检测中的应用研究

运用地质雷达的理论办法,了解某区域范围内已知埋藏物所造成的地质污染情况,以便为下一步布置工程工作和环境保护提供依据。这里还建立了用于区域污染识别的探地雷达探测数据处理模型,对地下污染情况进行了探测和识别,并与实际结果进行了比较。实践证明,该方法既可以提高探地雷达信号解释精度,又提高了工作效率。     

雷达测井数据处理方法及其应用效果

为了高效地进行数据处理,更加完善雷达测井数据处理,笔者提出一套雷达测井数据处理方法,自主研发了一套三天线雷达测井数据处理软件。通过对陕西省榆林市某矿井实测雷达测井数据进行坏道剔除、重采样、深度校正、直流消除、带通滤波、背景消除和信号增益等一系列处理,得到高信噪比的雷达剖面。结果表明：通过波速及走时等信息推算出该方法有效径向探测距离达到11 m,并且在天线三雷达剖面的410～415 m井段处可明显识别出破碎带异常,说明了该处理方法及流程的可行性。     

"甘肃平凉防雹减灾试验研究"简介

"甘肃平凉防雹减灾试验研究"是甘肃省政府资助的长达十年之久的试验项目,该试验研究始于1986年,其主要任务是平凉市的防雹减灾工作.一方面利用中国科学院兰州高原大气物理研究所70年代在平凉市防雹研究工作的成果指导防雹作业,一方面拿出少量的经费用于科研.研究工作主要集中在雹云的预报、预警、识别和作业技术方面.其关键技术有下面几点:     

南疆近60 a来冰雹灾害时空变化特征分析

冰雹是新疆主要灾害性天气之一, 利用新疆南疆各地州1949-2008年发生的冰雹频次及其造成的受灾农田面积、 经济损失、 死亡牲畜等资料, 分析了南疆冰雹频次时空变化特征及各县市雹灾灾情.结果表明: 南疆冰雹发生频次近60 a间呈明显上升趋势, 1950年代至1980年代中期发生频次较低, 1986年开始发生频次显著上升, 至1991年达最高值, 由1996年起又减少到2003年谷底.冰雹发生主要集中在3-10月, 其中6月频次最多, 年均3.6次, 其次5月为3.1次. 巴州、 阿克苏、 喀什和克州以春末到夏季型降雹为主, 和田以春季型降雹为主.冰雹一天之内多发生在白天到傍晚时段, 持续时间多为1~20min. 南疆冰雹空间分布上西北部多于东南部, 山脉的背风坡多于迎风坡, 山间盆地多于开阔平原.对各地州而言, 阿克苏冰雹发生频次最多, 属于严重雹灾区; 喀什和巴州次于阿克苏, 属于重雹灾区; 克州与和田冰雹发生频次较少, 属于中雹灾区. 冰雹受灾农田面积和经济损失随着工农业生产总值的增长而增加, 而且增长速度比经济增长速度快. 建议加大防雹投入力度, 建立人工防雹系统, 以减少雹灾损失, 遏制工农业发达地区雹灾损失严重和随着国民经济的增长雹灾损失增加的现象.     

基于雷达干涉失相干特性提取冰川边界方法研究

冰川边界提取是冰川变化监测的一个重要参数,光学冰川遥感方法可以有效、精确的提取出无云区域冰川的边界信息,但是,对于有云区域和天气状况不好的冰川区域却很难获取冰川的边界信息.提出了一种基于SAR干涉失相干特性的提取冰川边界信息的方法,该方法利用内陆冰川表面失相干问题严重的特点,通过比对冰面相干系数和非冰覆盖区域的相干系数,采用阈值法对冰川区域的干涉相干图进行冰川边界的提取.根据上述方法提取冰川边界的精度达到89%,结果表明利用雷达干涉失相干特性进行冰川边界提取是可行的.     

Radar research of hailstorm formation and development over the central part of Northern Caucasus (Russia). Organization and main results of the regional hail suppression projects

The results of the analysis of long-term radar research of the hailstorms over the central part of the Northern Caucasus are presented. Radar observations which formed the basis of experimental material were carried out continuously from the moment of the first radar echo registration until complete dissipation of the hailstorm. The computerized system of collecting, processing and the analysis of the radar information was used. Time discretization of the radar parameters of the full spatial observation averaged 3 min. Statistical sampling of radar data included 392 hail cells, for each of which the time distributions of the radar parameters both measured and calculated using the computerized system were compiled. Distribution of the hailstorms first radar echo formation zones over the observation region was compiled. Areas with the maximum frequency of the hailstorms first radar echo formation were defined. The hailstorm trajectories were analyzed. Four main types of hail cell trajectories were selected, which included 86 % of the supercells and 64 % of the long-lived multicells. The dynamic parameters of the hail core formation and development were analyzed. The hail storm characteristics of the Northern Caucasus are compared to the hail storms of Mendoza, Argentina, and Alberta, Canada. The bases of the organization of regional hail suppression services which use the rocket technology of cloud seeding are presented. The results of the cloud seeding operations during recent years are shown.     

大气重力波布阵探测灾害性冰雹过程的研究

1995－1997年在冰雹灾害频繁的贵州省中西部进行了灾害性冰雹过程的大气重力波布阵探测研究，结果表明使用动态谱分析技术的大气重力波三测点布探测方法不仅可以在冰雹灾害发生3－24小时以前获得可靠的预警信息，而且可连续跟踪监测波原所在方位，为冰雹过程发生发展的理论机制及预报方法和研究和人工防雹作业提供了一些新的线索和一种新的技术手段。     

Radar research of hailstorm formation and development over the central part of Northern Caucasus (Russia). Organization and main results of the regional hail suppression projects

The results of the analysis of long-term radar research of the hailstorms over the central part of the Northern Caucasus are presented. Radar observations which formed the basis of experimental material were carried out continuously from the moment of the first radar echo registration until complete dissipation of the hailstorm. The computerized system of collecting, processing and the analysis of the radar information was used. Time discretization of the radar parameters of the full spatial observation averaged 3 min. Statistical sampling of radar data included 392 hail cells, for each of which the time distributions of the radar parameters both measured and calculated using the computerized system were compiled. Distribution of the hailstorms first radar echo formation zones over the observation region was compiled. Areas with the maximum frequency of the hailstorms first radar echo formation were defined. The hailstorm trajectories were analyzed. Four main types of hail cell trajectories were selected, which included 86 % of the supercells and 64 % of the long-lived multicells. The dynamic parameters of the hail core formation and development were analyzed. The hail storm characteristics of the Northern Caucasus are compared to the hail storms of Mendoza, Argentina, and Alberta, Canada. The bases of the organization of regional hail suppression services which use the rocket technology of cloud seeding are presented. The results of the cloud seeding operations during recent years are shown.

     

Hailstorm Loss Modelling and Risk Assessment in the Sydney Region,Australia

This paper summarises work undertaken at the Natural Hazards ResearchCentre with regard to the compilation, climatological analysis and lossmodelling application of hailstorm data for the greater Sydney region. Itdiscusses the results from two associated research areas: ENSO-relatedclimatological analysis of hailstorm data, and insurance sector orientedrisk assessment of hail losses in the Sydney basin. The two complementaryprojects led to the development of HailAUS, the first hail loss model for amajor urban area in Australia. The hazard occurrence module of the hailloss model is based on a comprehensive, good quality hail data set, whoseinterannual variability has previously been analysed with regard to the ENSOsignal. The paper presents the major results of the local hail data climatology,and outlines the structure of the hail loss model. The important practicalapplications that result from the collection and analysis of good qualityclimatological data are noted.     

Landslide susceptibility mapping by using statistical analysis in the North Anatolian Fault Zone (NAFZ) on the northern part of Suşehri Town,Turkey

Numerical modelling techniques enable forcing function at the point of contact in a hail impact to be predicted, but many models available in commercial software have not been verified by experiments. A recently developed deterministic model is only able to simulate the impact action of an idealised spherical hailstone. Results recorded from the impact testing of non-spherical simulated hail ice specimens show wide scatters and are different to spherical specimens. A stochastic model has been developed in this study to incorporate the variability of the modelling parameters representing the behaviour of non-spherical ice specimens and to generate time histories of contact force for given impact scenarios. A probabilistic model which was developed previously to provide predictions of the distribution of the size of hailstones in a severe category hailstorm has also been further developed to provide predictions of the maximum contact force values in probabilistic terms. For any given pre-defined forcing function denting into the surface of a steel plate can be estimated by finite element analysis or experimentally on a test machine. The new knowledge base introduced in this paper can be used by manufacturers and designers to re-assess existing products and in making improvements to future installations taking into account cost–benefit considerations.     

Urban hailstorms: a view from Alberta

Urban hailstorms are rarely studied in detail. This work documents five urban storms in Alberta where damage has, on three occasions, set the record for Canada's most costly natural disaster. Information from newspapers, insurance companies, and disaster assistance programs was utilized to supplement meteorological records and information obtained from public surveys.The record-breaking hail swath which accompanied the 1987 Edmonton tornado was mapped using over 800 responses to an unprecedented newspaper survey. Tennis ball sized hail struck 125 km² of the city. Record-sized hailstones for Alberta were collected. Citizens' measurements of giant hailstones were compared to laboratory measurements. The rural storms were tracked using lightning detector information and damage was mapped using crop insurance and disaster assistance claims. The tornado-bearing storm was found to have a unique track.A late-season hailstorm which struck Calgary in 1991 was mapped using homeowner insurance claims organized by postal areas. Nine out of thirty areas had claims rates exceeding 50%, mainly for shingle replacement. Experiences of claims adjustors and an informal public survey were also utilized. Rural storms were mapped using weather radar and crop losses. The radar beam was strongly attenuated when it passed through hail-bearing storms and, thus, its ability to detect large hail was compromised.Weather conditions, urban and rural damaged areas, and insurance payments were compared for all five local hailstorms. These storms were discussed within the context of the long history of Alberta hail research and current trends in technology implementation. Forecasting of these hailstorms using conventional severe weather indicators was difficult in Calgary because of that city's proximity to the mountains. Hailstorms that struck Munich, Denver, and Toowoomba (Australia) were also discussed, and the hailstones collected from the great Munich storm were compared to those collected from the Edmonton storms.     

基于FY-3A遥感数据的冰岛火山灰云识别

2010年4月至5月期间冰岛艾雅法拉火山喷发造成了欧洲航空业史无前例的瘫痪以及巨大的经济损失,其严重影响再次显示,对火山灰云进行有效监测的重要性。火山灰云是由火山碎屑物及气体组成的混合物,火山碎屑物主要由直径小于2mm的岩石、矿物、火山玻璃碎片组成,火山灰云中的气体主要包括水汽、CO2、SO2、H2S、CH4、CO、HCL、HF、HBr、和NOx等。使用具有我国自主知识产权的FY-3A/VIRR数据,对此次艾雅法拉火山喷发的不同阶段选取具有典型风向变化的日期,采用分裂窗亮温差算法(SWTD)、RGB真彩色方法、中红外波段数据等进行火山灰云的识别,并将结果与冰岛地区的火山灰监测报告以及前人的研究结果进行对比研究,结果表明:火山喷发初期火山灰云中较高含量的水汽会补偿反面吸收的影响,妨碍分裂窗亮温差算法(SWTD)对火山灰云的识别,而中红外波段数据因对高温物体的敏感性,不受水汽的影响,对喷发初期较高温度的火山灰云识别效果较好;在喷发中期,火山灰云浓度较大时三种方法均表现良好,卫星图像中火山灰云的位置信息及漂移方向均非常清晰,且同气象条件相吻合,验证了识别方法的正确性。该项结果表明,具有我国自主知识产权的FY-3A数据能够达到监测火山灰云的目的,而如何更加清晰地界定火山灰云的边界位置以及更加准确的计算出火山灰云的浓度需要进一步的深入研究。     

基于SBAS-InSAR技术的深切割高山峡谷区滑坡灾害早期识别

近年来,高山峡谷区滑坡灾害频频发生,给人民生命和财产安全带来严重威胁。针对多数学者利用SAR单轨道数据对高山峡谷区滑坡进行早期识别,存在SAR成像几何畸变造成部分滑坡不能识别、识别结果不全面等问题。为全面准确的对高山峡谷区滑坡隐患进行早期识别,文章采用SBAS-InSAR技术,以东川小江沿线两侧深切割高山峡谷区为研究区,通过升降轨SAR数据结合互补的方式进行滑坡灾害隐患识别,引入高分辨率光学影像等作为辅助识别,最终共识别出18处滑坡灾害体,其中5处为高风险潜在滑坡,并对三类典型潜在滑坡进行分析。分析结果表明:利用升降轨SAR数据结合互补的方式,能有效避免SAR单轨道数据在高山峡谷地区产生的几何畸变问题,同时,该方法能更为准确全面地对高山峡谷区滑坡隐患进行早期识别,为防灾减灾事业及政府部门决策提供一种有效的手段。     

基于超前钻探测试的隧道地层智能识别方法

可靠地识别掌子面前方地层是保证隧道工程稳定与安全的重要因素之一。传统的超前地质预报方法不能同时保证有高识别精度、低实施成本和占用少的施工时间,对于不同地质情况的地层识别通用性不强。在传统超前钻孔的同时获取掌子面前方围岩钻探测试数据,实时获取不同深度岩层情况,将大大提高超前预报效率,方便快捷,不影响施工,但目前缺乏客观、准确的地层识别方法。提出了一种基于神经网络的钻探测试数据智能分析和地层识别方法,对楚大高速公路九顶山隧道超前钻探测试数据进行了深入分析,通过隧道开挖后所揭示地层对分析方法进行了验证。结果表明：单一钻进参数用于地层识别的错误率在35%左右,打击能和打击数、送水压力和送水流量的参数组合不能显著提升地层识别准确率;钻进速度、扭矩、回转数、推进力的参数组合可降低地层识别错误率至22%。在神经网络模型中引入钻进参数的标准差,可大幅降低错误率,可使地层划分错误率下降9%～12%;多参数组合下的神经网络钻探测试神经网络模型对随机抽样的地层识别错误率小于10%,对单个钻孔的地层识别错误率小于14%。