Landslide susceptibility zonation method based on C5.0 decision tree and K-means cluster algorithms to improve the efficiency of risk management

Machine learning algorithms are an important measure with which to perform landslide susceptibility assessments,but most studies use GIS-based classification methods to conduct susceptibility zonation.This study presents a machine learning approach based on the C5.0 decision tree(DT)model and the K-means cluster algorithm to produce a regional landslide susceptibility map.Yanchang County,a typical landslide-prone area located in northwestern China,was taken as the area of interest to introduce the proposed application procedure.A landslide inventory containing 82 landslides was prepared and subse-quently randomly partitioned into two subsets:training data(70％landslide pixels)and validation data(30％landslide pixels).Fourteen landslide influencing factors were considered in the input dataset and were used to calculate the landslide occurrence probability based on the C5.0 decision tree model.Susceptibility zonation was implemented according to the cut-off values calculated by the K-means clus-ter algorithm.The validation results of the model performance analysis showed that the AUC(area under the receiver operating characteristic(ROC)curve)of the proposed model was the highest,reaching 0.88,compared with traditional models(support vector machine(SVM)=0.85,Bayesian network(BN)=0.81,frequency ratio(FR)=0.75,weight of evidence(WOE)=0.76).The landslide frequency ratio and fre-quency density of the high susceptibility zones were 6.76/km2 and 0.88/km2,respectively,which were much higher than those of the low susceptibility zones.The top 20％interval of landslide occurrence probability contained 89％of the historical landslides but only accounted for 10.3％of the total area.Our results indicate that the distribution of high susceptibility zones was more focused without contain-ing more"stable"pixels.Therefore,the obtained susceptibility map is suitable for application to landslide risk management practices.     

毫米波雷达云回波的自动分类技术研究

毫米波雷达在云探测方面比厘米波天气雷达和激光雷达具有显著优势,可获得更多的云粒子信息,是研究云特性的主要遥感探测设备。为了开展对毫米波雷达探测的云回波进行自动分类的研究,利用161次云回波的个例数据,统计得到了卷云、高层云、高积云、层云、层积云和积云6类云型的特征量和其他参量的数值范围,利用分级的多参数阈值判别方法,达到了自动分类的目标,通过与人工分类的初步验证,两种分类结果的一致性达到84%,其中,层云和积云的识别一致较低的原因在于样本数据有限,仅有6次层云和8次积云的个例样本数据。通过更多样本的处理,提取的特征参量更可靠,自动分类的准确率会得到提高,以便将基于毫米波雷达的云分类技术应用于将来的云观测自动化业务。      

On the linkage between changes in cloud cover and precipitation extremes over Central India

In this study, linkage between changing characteristics of precipitation extremes and cloud covers over Central India is explored during summer monsoon period using Satellite data (1998–2015). This is a first attempt to relate the changes in cloud cover to the changes in precipitation extremes. Non-rainy cirrus clouds are excluded from this study. Results show that heavy rainfall (≥ 60 mm/day) is associated with cold cloud tops (Tb≤220 K) while moderate rainfall (<60 mm/day and ≥20 mm) occurs mostly with middle clouds (Tb>220 K and ≤245 K). Low level clouds (Tb> 245 K) are responsible for light rainfall (<20 mm/day). Increases in top 20%, 10%, 5% and 1% heavy precipitation relate well with the increases in very deep convective, deep convective and convective cloud cover. Among these relations, increase in top 5% heavy precipitation relates best with increase in very deep convective cloud cover. Decrease in bottom 30% low precipitation relates with decrease in low level cloud cover. The results reported in this study fit into the framework of how weather extremes respond to climate change.     

基于DBSCAN与网格搜索的雷电定位算法

定位精度是评价雷电定位网络的重要指标之一,定位算法直接影响雷电探测结果的精度。雷电监测系统探测数据误差不可避免,传统定位算法不具备抗误差干扰能力,迭代计算易发散,定位结果精度不高。为了满足实际应用需求,提出一种新的雷电定位算法DG-LLA（DBSCAN and grid-search lighting location algorithm）,在定位计算中引入DBSCAN（density-based spatial clustering of applications with noise）方法与网格搜索方法。通过仿真与国家雷电监测网实际定位结果对比分析定位算法性能。结果表明:到达时间差（time difference of arrival,TDOA）法和Taylor级数展开法定位误差较大,仿真区域的均方根误差分别为982 m和668 m;定位中引入DBSCAN方法后,均方根误差明显减小为406 m,引入DBSCAN方法和网格搜索方法后,均方根误差减小为349 m;在相同回击数据条件下,算法DG-LLA与国家雷电监测网相比定位数量更多,回击数据的利用率从43.4%提升到51.5%,新增定位结果周围雷达回波特征较强,定位精度更高。     

Estimation of flow enhancement in the Kinneret watershed due to cloud seeding

A comparison is carried out between historical records of the flow measured in Kinneret watershed during and prior to the time of cloud seeding for rainfall enhancement. Precipitation series for the control area of the meteorological experimentation serve as a reference for the comparison. The fluctuations of the flow, which would have occurred unless the effect of the seeding, are estimated by a linear regression on the precipitation as the control. The regression parameters are calibrated separately for the unseeded and for the seeded time series. The model with the parameters calibrated for the unseeded series is applied on the rainfall recorded during the seeded time, and vice versa. The difference between the measured and the computed data is attributed to the effect of cloud seeding. Similar comparisons are carried out with respect to rainfall series recorded at the target area and at the edge of the enhanced area.The results indicate that the flow from the affected sector of the watershed has been enhanced, with respect to the control, by 31×10⁶ m ³/year, at a significance level of 31. This enhancement is 5% of the volume which is generated in that area. The rates found with respect to the rainfall at the edge are higher than those found with respect to the control, while those with respect to the rainfall at the center of the target area are lower.     

中国水质站网规划中几个问题的探讨

在完成中国水质站网规划工作的基础上,对下列问题作了进一步探讨:(1)水质站网规划单元的划分;(2)聚类分析在水质站网规划中的应用;(3)站网密度与人口和经济发展的关系;(4)中国水质站网的分区特征。     

云、雨水酸度和离子浓度与其微物理参数的关系

本文1987年5-6月和1989年5-6月庐山地区的实测资料,分析了云、雨水酸度和离子浓度与其微物理量间的关系。结果表明,云、雨滴大小、含水量、谱宽等微物理量的演变直接影响着其酸度和离子浓度。     

华南铀矿省铀矿构造成矿作用及其时空演化

华南铀矿省内已发现的工业铀矿床类型有花岗岩型、火山岩型、碳硅泥岩型、碳酸盐岩型和砂岩型。在空间上。铀矿床受环形及线性构造控制,并受上地幔构造制约。幔凸、幔凹阳地幔变异带互相交错,造成本区地幔的极不均一性,这是本矿省成矿的深部地质构造背景。通常铀矿床呈面型成环状分布,大型环形构造控制铀矿省,中型和小型环形构造控制铀矿田和矿床的产出。     

局地强风暴云团与中尺度锋面波动

本文分析了1982—1988年2—5月份的41例与中尺度锋面波动相关联的局地强风暴云团的活动和演变特征，给出了其强盛期天气系统配置模式及其在天气尺度云系演变中的表现型式，最后对其移动做了一般性讨论。