A spatial cross‐correlation model of spectral accelerations at multiple periods

Many seismic loss problems (such as disruption of distributed infrastructure and losses to portfolios of structures) are dependent upon the regional distribution of ground‐motion intensity, rather than intensity at only a single site. Quantifying ground‐motion over a spatially‐distributed region therefore requires information on the correlation between the ground‐motion intensities at different sites during a single event. The focus of the present study is to assess the spatial correlation between ground‐motion spectral accelerations at different periods. Ground motions from eight well‐recorded earthquakes were used to study the spatial correlations. On the basis of obtained empirical correlation estimates, we propose a geostatistics‐based method to formulate a predictive model that is suitable for simulation of spectral accelerations at multiple sites and multiple periods, in the case of crustal earthquakes in active seismic regions. While the calibration of this model and investigation of its implications were somewhat complex, the model itself is very simple to use for making correlation predictions. A user only needs to evaluate a simple equation relying on three sets of coefficients provided here to compute a correlation coefficient for spectral values at two periods and at a specified separation distance. These results may then be used in evaluating the seismic risk of portfolios of structures with differing fundamental periods. Copyright © 2012 John Wiley & Sons, Ltd.     

Fully probabilistic seismic displacement analysis of spatially distributed slopes using spatially correlated vector intensity measures

Earthquake‐induced slope displacement is an important parameter for safety evaluation and earthquake design of slope systems. Traditional probabilistic seismic hazard analysis usually focuses on evaluating slope displacement at a particular location, and it is not suitable for spatially distributed slopes over a large region. This study proposes a computationally efficient framework for fully probabilistic seismic displacement analysis of spatially distributed slope systems using spatially correlated vector intensity measures (IMs). First, a spatial cross‐correlation model for three key ground motion IMs, that is, peak ground acceleration (PGA), Arias intensity, and peak ground velocity, is developed using 2686 ground motion recordings from 11 recent earthquakes. To reduce the computational cost, Monte Carlo simulation and data reduction techniques are utilized to generate spatially correlated random fields for the vector IMs. The slope displacement hazards over the region are further quantified using empirical predictive equations. Finally, an illustrative example is presented to highlight the importance of the spatial correlation and the advantage of using spatially correlated vector IMs in seismic hazard analysis of spatially distributed slopes. Copyright © 2013 John Wiley & Sons, Ltd.     

Extracting water-related features using reflectance data and principal component analysis of Landsat images

This study aimed to map water features using a Landsat image rather than traditional land cover. We involved the original bands, spectral indices and principal components (PCs) of a principal component analysis (PCA) as input data, and performed random forest (RF) and support vector machine (SVM) classification with water, saturated soil and non-water categories. The aim was to compare the efficiency of the results based on various input data. Original bands provided 93% overall accuracy (OA) and bands 4–5–7 were the most informative in this analysis. Except for MNDWI (modified normalized differenced water index, with 98% OA), the performance of all water indices was between 60 and 70% (OA). The PCA-based approach conducted on the original bands resulted in the most accurate identification of all classes (with only 1% error in the case of water bodies). We therefore show that both water bodies and saturated soils can be identified successfully using this approach.     

Dependence of correlations between spectral accelerations at multiple periods on magnitude and distance

In this paper, the dependence of correlations between spectral accelerations at multiple periods on magnitude (M) and distance (R) has been investigated. For this purpose, a relatively large dataset of ground motion records (GMRs), containing 1551 records with a wide range of seismic characteristics, was selected. It is shown that the difference in the correlation coefficient is statistically meaningful when the general GMR dataset is divided into two subsets based on an arbitrary M or R. The observed difference is more meaningful in the case of magnitude when compared with distance. The general dataset of GMRs was then divided into four separate subsets based on optimum values of M and R, so that the four obtained subsets were given the greatest dissimilarity in terms of the correlation coefficients. The correlation coefficients between spectral accelerations at multiple periods were calculated in the case of the four subsets and compared with the available correlations in the literature. The conditional mean spectrum was also calculated by means of the conventional correlation coefficients, as well as by using the proposed M and R dependent correlation coefficients. The results show that, despite the commonly available findings in the literature, this dependence is significant and should not be neglected in the conditional spectra calculation process. Copyright © 2013 John Wiley & Sons, Ltd.     

Correlation model for spatially distributed ground‐motion intensities

Risk assessment of spatially distributed building portfolios or infrastructure systems requires quantification of the joint occurrence of ground‐motion intensities at several sites, during the same earthquake. The ground‐motion models that are used for site‐specific hazard analysis do not provide information on the spatial correlation between ground‐motion intensities, which is required for the joint prediction of intensities at multiple sites. Moreover, researchers who have previously computed these correlations using observed ground‐motion recordings differ in their estimates of spatial correlation. In this paper, ground motions observed during seven past earthquakes are used to estimate correlations between spatially distributed spectral accelerations at various spectral periods. Geostatistical tools are used to quantify and express the observed correlations in a standard format. The estimated correlation model is also compared with previously published results, and apparent discrepancies among the previous results are explained. The analysis shows that the spatial correlation reduces with increasing separation between the sites of interest. The rate of decay of correlation typically decreases with increasing spectral acceleration period. At periods longer than 2 s, the correlations were similar for all the earthquake ground motions considered. At shorter periods, however, the correlations were found to be related to the local‐site conditions (as indicated by site V_s30 values) at the ground‐motion recording stations. The research work also investigates the assumption of isotropy used in developing the spatial correlation models. It is seen using the Northridge and Chi‐Chi earthquake time histories that the isotropy assumption is reasonable at both long and short periods. Based on the factors identified as influencing the spatial correlation, a model is developed that can be used to select appropriate correlation estimates for use in practical risk assessment problems. Copyright © 2009 John Wiley & Sons, Ltd.     

Interbuilding interpolation of peak seismic response using spatially correlated demand parameters

Seismic structural responses recorded in instrumented buildings during an earthquake are used to provide insights into the demands placed on neighboring, noninstrumented buildings, using a framework to interpolate structural response demands across buildings. The interbuilding interpolation model relies on the spatial and structural correlations of responses in coregionally located buildings subjected to a seismic event. A dataset of response demands for a portfolio of reinforced concrete moment frame buildings is generated by performing nonlinear response history analyses on structural models using ground motions recorded from historical scenario earthquakes. The dataset is used to characterize the correlation between seismic demands across different buildings. Semivariograms are used to model spatial and structural correlations and then incorporated into a kriging algorithm, which forms the basis of the interpolation models. The effect of several model and dataset attributes and assumptions, for example, using intensity‐measure‐based versus engineering demand parameter‐based semivariograms, and size of training dataset relative to portfolio, on the overall performance are evaluated along with the limitations of the proposed model.     

主成分分析法在2013年灯塔M_S 5.1地震预测中的应用

以2013年1月23日灯塔M_S 5.1地震为研究对象,计算地震发生区域的地震频次N、b值、η值、A(b)值、M_f值、AC值、C值、D值、R_m值、能量E、响应比Y等11个地震活动性参数,并应用主成分分析方法,进行参数约简和数据挖掘,得到综合参数W,发现W在灯塔地震发生前有明显异常变化,说明主成分分析法在地震活动性参数的信息再挖掘中可以发挥巨大作用。     

利用主成分分析法研究余干地震台接收函数

对模型合成的理论接收函数进行主成分分析,研究在倾斜界面和各向异性影响下主成分形态变化规律,分析倾斜界面倾角和各向异性强度对接收函数R分量主成分贡献率的影响。对江西余干地震台的接收函数R分量主成分进行分析,结果发现:Ps震相的平均走时为3.4 s;台站下方介质的各向异性和倾斜界面同时存在,各向异性快轴大致呈NE向,倾斜界面倾向约170°。     

Processes governing river water quality identified by principal component analysis

The present study demonstrates the usefulness of principal component analysis in condensing and interpreting multivariate time‐series of water quality data. In a case study the water quality system of the lock‐regulated part of the River Neckar (Germany) was analysed, with special emphasis on the oxygen budget. Pooled data of ten water quality parameters and discharge, which had been determined at six stations along a 200 km reach of the river between the years 1993 and 1998, were subjected to principal component analysis. The analysis yielded four stable principal components, explaining 72% of the total variance of the 11 parameters. The four components could be interpreted confidently in terms of underlying processes: biological activity, dilution by high discharge, seasonal effects and the influence of wastewater. From analysing the data of single stations separately, these processes were found to be active throughout the complete reach. Considering the oxygen budget of the river, the variance of biological activity, representing the counteracting processes of primary production and microbial degradation, was found to be most important. This principal component explained 79% of the observed variance of oxygen saturation. In contrast, the analysis of a reduced data set from the 1970s showed that oxygen saturation was then dominated by discharge and temperature variations. The findings indicate that the oxygen budget used to be governed directly by the emission of degradable matter, whereas nowadays eutrophication is most important for extreme oxygen concentrations. Therefore, controlling eutrophication has to be the primary goal, in order to mitigate the rare episodes of pronounced oxygen over‐ and undersaturation in the future. Copyright © 2002 John Wiley & Sons, Ltd.     

Frequency‐dependent principal component analysis of multicomponent earthquake ground motions

Stochastic ground motion simulation techniques are becoming increasingly popular because of enhanced computation power enabling direct simulation of complex response quantities. Priestley process assumption is the most general approach for stochastic modeling of earthquake ground motion. However, a framework for multicomponent ground motion simulation using the general Priestley process assumption is not available. Multicomponent motions are useful especially when the correlation structure between them significantly influences the response. The present study proposes a framework for frequency‐dependent principal component analysis (PCA), which facilitates Priestley process–based simulation of multicomponent ground motions. The study focuses only on the frequency‐dependent PCA part, and the results show high dependency of the principal components/directions on the frequency bands of the signals. The present work also advocates that the frequency‐dependent PCA should be preferred to the conventional PCA as the former can address the issues related to the frequency‐independent uniform modulation associated with the latter.     

Stochastic simulation of regionalized ground motions using wavelet packets and cokriging analysis

Performance‐based earthquake engineering often requires ground‐motion time‐history analyses to be performed, but very often, ground motions are not recorded at the location being analyzed. The present study is among the first attempt to stochastically simulate spatially distributed ground motions over a region using wavelet packets and cokriging analysis. First, we characterize the time and frequency properties of ground motions using the wavelet packet analysis. The spatial cross‐correlations of wavelet packet parameters are determined through geostatistical analysis of regionalized ground‐motion data from the Northridge and Chi‐Chi earthquakes. It is observed that the spatial cross‐correlations of wavelet packet parameters are closely related to regional site conditions. Furthermore, using the developed spatial cross‐correlation model and the cokriging technique, wavelet packet parameters at unmeasured locations can be best estimated, and regionalized ground‐motion time histories can be synthesized. Case studies and blind tests using data from the Northridge and Chi‐Chi earthquakes demonstrate that the simulated ground motions generally agree well with the actual recorded data. The proposed method can be used to stochastically simulate regionalized ground motions for time‐history analyses of distributed infrastructure and has important applications in regional‐scale hazard analysis and loss estimation. Copyright © 2014 John Wiley & Sons, Ltd.     

Using principal component analysis in the investigation of groundwater hydrochemistry of Upper Jezireh Basin,Syria

Principal component analysis (PCA) was applied to hydrochemical and isotopic data of 34 groundwater samples. This allowed the reduction of 20 variables to four significant PCs that explain 81.9% of the total variance; F1 (47.1%) explains the groundwater mineralization, whereas F2 (17%) shows isotopic enrichment and nitrate pollution. Based on an iso-factor scores map of F1, three water zones were delineated: Zone A (F1 < ?1), with fresh groundwater from the unconfined aquifer; Zone B (1 > F1 > ?1), with moderate mineralization from the confined–unconfined aquifer boundary; and Zone C (F1 > 1), with the most mineralized hot water from the confined aquifer. The iso-factor scores map of F2 delineates positive values representing samples from the unconfined aquifer, with freshwater and nitrate contamination associated with stable isotope enrichment, whereas negative values represent samples from the confined aquifer. The results clearly demonstrate the usefulness of PCA in groundwater hydrochemistry investigations.     

基于核主成分分析的时间域航空电磁去噪方法

时间域航空电磁数据往往在测量过程中受到天然和人文噪声的干扰.如果不能很好滤除这些电磁噪声,那么将会降低资料质量、影响反演的精度,甚至获得错误的解释结果.本文提出了一种基于核主成分分析的去噪方法,通过核主成分分析提取叠加后衰减曲线的主成分,然后使用能量占比方法分离反映地下介质的有效信号和噪声,最后使用反映地下介质的特定成分进行重构.本文所推荐的去噪方法不仅能剔除天然噪声,例如天电产生的尖脉冲或者振荡,而且能有效地抑制人文噪声.分别使用基于核主成分分析的去噪方法,以及AeroTEM软件的处理方法对同样的吊舱式时间域直升机航空电磁勘查系统实测数据进行处理,并比较其结果.处理结果表明：所推荐的去噪方法要优于AeroTEM软件.     

基于主成分降维的总悬浮物浓度遥感估算模型适用性分析

总悬浮物浓度(C_TSM)是水质评价的重要参数.为了提高内陆Ⅱ类水体总悬浮物浓度估算的精度,利用主成分分析方法对2009年4月太湖水体实测高光谱数据进行降维处理,进而以不同数量的主成分作为变量,分别构建总悬浮颗粒物浓度的多元线性回归估算模型并比较这些模型的效果,从而确定最优的主分量个数;结合近年运行的高光谱传感器,对模型的适用性进行评价.结果表明:①前三个主成分(PC₁、PC₂、PC₃)从不同侧面涵盖了悬浮物浓度信息,它们与ln(C_TSM)的相关系数分别为0.728、0.401和0.403;②当主成分个数为6时,模型达到最优;模型的精度高于4个传统经验模型;③在400~850 nm之间,波段数大于45的高光谱传感器数据都能利用主成分分析的方法构建精度较高的总悬浮物浓度估算模型;此外,MERIS、HJ1-HSI、Hyperion和CHRIS这些常用的高光谱传感器的波段设置,都适合于主成分建模.     

基于独立分量分析的多次波自适应相减技术

针对多次波自适应相减这个关键问题,文中首次提出利用独立分量分析技术来实现多次波和一次波的分离（简称ICAAMS）. 现有的多次波自适应相减技术大都是采用输出信号（一次波）能量最小准则,基于二阶统计量的技术. 本文提出的ICAAMS采用了输出信号非高斯性最大准则,并利用高阶统计量来表征非高斯性. 简单的褶积模型和复杂的有限差分模型资料处理结果表明, 本文提出的方法可以有效地分离一次波和多次波.     

华北地区地震活动的连续小波时频演化特征研究

利用Morlet小波变换对华北地区的地震活动进行了小波动态周期分析。结果分析表明，地震活动既存在着比较稳定的周期，也具有一定时变性。华北地区1000年来，特别是1500年以来存在着比较稳定的300年周期和60年左右周期，还有一些不断随时间变化的周期现象，不断地产生和消失。小波分析揭示的历史地震活动的周期时变性，一方面提示我们，在依据地震活动周期规律外推地震趋势时，要应用近期周期分析结果和使用比较稳定的周期成份；另一方面也告诉我们，短周期突变的不可预知性对相对短趋势预测的结果准确度带来很大影响。     

独立分量分析及其在地震信息处理中应用初探

独立分量分析(ICA)是最近才发展起来的一种统计学方法，旨在寻求对非高斯分布数据进行有效表示，使得各个分量在统计学上独立，或者尽最大可能地独立。许多应用中，这种表示意在获取数据的基本结构，包括特征提取和信号分离。本文给出ICA的基本理论和快速算法，并对FastICA稍作改进，在分析地震信号特点的基础上，对其在地震信息处理中应用进行初步探索，表明ICA在地震信号处理中具有应用前景。     

Identification and classification of transient pulses observed in magnetometer array data by time-domain principal component analysis filtering

A method for identification of pulsations in time series of magnetic field data which are simultaneously present in multiple channels of data at one or more sensor locations is described. Candidate pulsations of interest are first identified in geomagnetic time series by inspection.Time series of these ‘‘training events' ' are represented in matrix form and transpose-multiplied to generate timedomain covariance matrices. The ranked eigenvectors of this matrix are stored as a feature of the pulsation. In the second stage of the algorithm, a sliding window(approximately the width of the training event) is moved across the vector-valued time-series comprising the channels on which the training event was observed. At each window position, the data covariance matrix and associated eigenvectors are calculated. We compare the orientation of the dominant eigenvectors of the training data to those from the windowed data and flag windows where the dominant eigenvectors directions are similar. This was successful in automatically identifying pulses which share polarization and appear to be from the same source process. We apply the method to a case study of continuously sampled(50 Hz) data from six observatories, each equipped with threecomponent induction coil magnetometers. We examine a90-day interval of data associated with a cluster of four observatories located within 50 km of Napa, California,together with two remote reference stations-one 100 km to the north of the cluster and the other 350 km south. When the training data contains signals present in the remote reference observatories, we are reliably able to identify and extract global geomagnetic signals such as solar-generatednoise. When training data contains pulsations only observed in the cluster of local observatories, we identify several types of non-plane wave signals having similar polarization.