基于逻辑回归的四川青川县区域滑坡灾害预警模型

四川省青川县滑坡灾害群发,点多面广,区域滑坡灾害预警是有效防灾减灾的重要手段,预警模型是成功预警的核心。由于研究区滑坡诱发机理复杂、调查监测大数据及分析方法不足等原因,传统区域地质灾害预警模型存在预警精度有限、精细化不足等问题。文章在青川县地质灾害调查监测和降水监测成果集成整理与数据清洗基础上,构建了青川县区域滑坡灾害训练样本集,样本集包括地质环境、降雨等27个输入特征属性和1个输出特征属性,涵盖了青川县近9年（2010—2018年）全部样本,数量达1 826个（其中,正样本613个,负样本1 213个）。基于逻辑回归算法,对样本集进行5折交叉验证学习训练,采用贝叶斯优化算法进行模型优化,采用精确度、ROC曲线和AUC值等指标校验模型准确度和模型泛化能力。其中,ROC曲线也称为“受试者工作特征”曲线;AUC值表示ROC曲线下的面积。校验结果显示,基于逻辑回归算法的模型训练结果准确率和泛化能力均较好（准确率94.3%,AUC为0.980）。开展区域滑坡实际预警时,按训练样本特征属性格式,输入研究区各预警单元27个特征属性,调用预先学习训练好的模型,输出滑坡灾害发生概率,根据输出概率分段确定滑坡灾害预警等级。当输出概率P≥40%且P<60%时,发布黄色预警;当输出概率P≥60%且P<80%时,发布橙色预警;当输出概率P≥80%时,发布红色预警。     

滑坡对降雨响应的多指标监测及综合预警探析:以赣南罗坳滑坡为例

降雨量和位移是当前降雨型滑坡监测预警最常用的指标。然而,降雨量和位移监测结果只能反映降雨作用下滑坡的变形情况,不能揭示滑坡内在物理力学性状对降雨的响应。因此,除降雨量和位移监测之外,建立包括体积含水率、基质吸力等反映滑坡动态演化过程的关键指标监测体系必将成为今后更真实地把握滑坡内在演化趋势、更准确地建立滑坡综合预警判据的最有效手段。笔者对赣南地区典型降雨型滑坡进行了多指标监测及综合预警示范研究。结果表明:(1)在降雨条件下滑坡土体内部体积含水率、基质吸力和温度等多指标均产生有规律的动态响应;(2)随着降雨的持续,滑体体积含水率与基质吸力的变化均具有显著的滞后现象;(3)体积含水率和基质吸力变化速率与滑体位移具有显著的正相关性;(4)滑体温度分布变化规律受大气温度和体积含水率的共同影响。以实测数据的滑坡稳定性分析为基准,在考虑实际降雨入渗深度与滑坡稳定性的关联度上,建立了包括日降雨量、体积含水率增加速率、基质吸力减小速率以及位移速度多元指标预警方法体系,提出了基于关键指标综合预警体系及确定方法,旨在为降雨滑坡准确预警提供新模式。     

1822号“山竹”台风龙卷过程观测与预警分析

2018年9月17日09:37—10:00,在登陆台风“山竹”外围螺旋雨带中,广东省佛山市三水区到肇庆市四会区发生了EF2级强龙卷,龙卷路径长度18 km,持续时间23 min,平均时速47 km/h,导致不少建筑物损毁。本次过程佛山市进行了龙卷预警试验,提前37 min发布了龙卷预警,龙卷没有造成人员伤亡。利用观测资料对产生强龙卷的环境场特征、地面自动站和雷达观测的中小尺度特征以及龙卷预警试验进行了分析,结果表明:产生龙卷的微型超级单体出现在台风外围和副高边缘之间的强东南急流中,具有低层辐合、高层辐散、水汽充足等典型台风外围龙卷环流形势特征;强的低空0~1 km垂直风切变、大的风暴相对螺旋度和低的抬升凝结高度等环境条件利于龙卷的生成;龙卷影响时,邻近地面自动气象站观测要素表现出明显的信号,瞬时大风和最低气压的极值区呈东南至西北向带状分布,与龙卷路径一致,龙卷过境前后,单站气压“漏斗”明显,5 min降压/升压幅度达-2.5 hPa/+2.1 hPa;广州多普勒天气雷达探测到龙卷母体风暴的低层钩状回波和入流缺口特征,以及低层强中气旋和类TVS特征。预警试验初步表明,对台风龙卷高发区,在环境场有利情况下,若低层出现中等或以上强度中气旋,其底高在1 km以下,可以考虑发布龙卷预警。      

次声波在海啸灾害预警中的应用探讨

介绍了次声波的基本物理概念和次声波的传播特性,概述了次声波的研究现状,提出了利用次声波来预警海啸灾害的建议,并对其可行性进行了分析。     

太湖蓝藻水华遥感动态监测预警信息系统

太湖蓝藻水华暴发对其周边地区的生态、经济和社会造成了严重影响,对蓝藻水华进行监测预警对于改善太湖水环境具有重要意义。本文在分析蓝藻水华遥感监测预警研究动态基础上,从信息系统构建角度,结合太湖蓝藻水华监测情况,对太湖蓝藻水华遥感监测预警流程进行系统性分析,提出建立太湖蓝藻水华遥感监测预警系统的总体框架,利用C#语言和IDL语言实现了该框架的软件架构,并且论述了实现该系统框架所采用的系统集成技术和软件模式技术。     

Cartograms for Use in Forecasting Weather-Driven Natural Hazards

This paper evaluates the potential of using cartograms for visualizing and interpreting forecasts of weather-driven natural hazards in the context of global weather forecasting and early warning systems. The use of cartograms is intended to supplement traditional cartographic representations of the hazards in order to highlight the severity of an upcoming event. Cartogrammetric transformations are applied to forecasts of floods, heatwaves, windstorms and snowstorms taken from the European Centre for Medium-range Weather Forecasts (ECMWF) forecast archive. Key cartogram design principles in standard weather forecast visualization are tested. Optimal cartogram transformation is found to be dependent on geographical features (such as coastlines) and forecast features (such as snowstorm intensity). For highly spatially autocorrelated weather variables used in analysing several upcoming natural hazards such as 2m temperature anomaly, the visualization of the distortion provides a promising addition to standard forecast visualizations for highlighting upcoming weather-driven natural hazards.     

地震预警断层参数实时识别方法(FinDer)详解及其性能初步评价

详细阐述了美国加州ShakeAlert地震预警系统中使用的断层破裂参数实时测定方法(FinDer FinDer:Finite Fault Rupture Detector)的技术细节.利用1999年集集Mw7.6级地震、2008年汶川Ms8.0级地震及2013年芦山Ms7.0级地震,以及日本K-NET台网记录的80次中...     

Satellite-derived rainfall thresholds for landslide early warning in Bogowonto Catchment,Central Java,Indonesia

Satellite rainfall products for landslide early warning prediction have been spotlighted by several researchers, in the last couple of decades. This study investigates the use of TRMM and ERA-Interim data, for the determination of rainfall thresholds and the prediction of precipitation, respectively, to be used for landslide early warning purposes at the Bogowonto catchment, Central Java, Indonesia. A landslide inventory of 218 landslides for the period of 2003–2016 was compiled, and rainfall data were retrieved for the landslide locations, as given by 6 ground stations, TRMM, and ERA-Interim data. First, rainfall data from the three different sources was compared in terms of correlation and extreme precipitation indices. Second, a procedure for the calculation of rainfall thresholds for landslide occurrence was followed consisting of four steps: i) the TRMM-based rainfall data was reconstructed for selected dates and locations characterized by landslide occurrence and non-occurrence; ii) the antecedent daily rainfall was calculated for 3, 5, 10, 15, 20 and 30 days for the selected dates and locations; iii) two-parameter daily rainfall-antecedent rainfall thresholds were calculated for the aforementioned dates; after analysis of the curves the optimum number of antecedent rainfall days was selected; and (iv) empirical rainfall thresholds for landslide occurrence were determined. The procedure was repeated for the entire landslide dataset, differentiating between forested and built-up areas, and between landslide occurrence in four temporal periods, in relation to the monsoon. The results indicated that TRMM performs well for the detection of very heavy precipitation and can be used to indicate the extreme rainfall events that trigger landslides. On the contrary, as ERA-Interim failed to detect those events, its applicability for LEWS remains limited. The 15-day antecedent rainfall was indicated to mostly affect the landslide occurrence in the area. The rainfall thresholds vary for forested and built-up areas, as well as for the beginning, middle and end of the rainy season.     

我国海啸灾害及预警现状与建议

本文给出了海啸定义及其它有关概念与术语,简要给出了我国及全球的海啸灾害.全面介绍了我国海啸监测能力、预警现状、研究成果以及有关国际合作的情况,重点介绍了我国自主研制的海啸数值模式,利用该模式我们模拟了2004年12月26日发生在印度洋大海啸以及假想的发生在我国南海的海啸对周边国家的影响.对海啸预警中存在的问题及未来急需开展的工作,作者也将一一给予阐述.     

基于KiK-net数据的偏振分析方法计算震中方位角影响因素分析

基于日本KiK-net强震动观测记录,以提高计算震中方位角的准确性为目标,研究记录参量(加速度、速度、位移)、计算时间窗和滤波频带对偏振分析方法计算震中方位角结果的影响。结果表明,在该研究的地震数据条件下,采用位移记录、计算时间窗为1s、滤波频带取0.1～20Hz时,可以获得最佳的震中方位角计算结果,震中方位角的计算偏差为45°时所占比例为88%。