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.     

基于独立分量分析的南极半岛GNSS网区域滤波

高精度GNSS速度场是研究地壳垂向运动及板块运动的基础,能够为冰川均衡调整（Glacial Isostatic Adjustment,GIA）的建模提供外部检核和新的约束.共性误差（Common Mode Error,CME）是区域连续GNSS时间序列中存在的一种与时空相关的主要误差源,通过空间滤波可有效的降低共性误差的影响,提高坐标时间序列的精度.目前广泛采用的主分量分析法（Principal Component Analysis,PCA）,基于二阶统计量（方差和协方差）进行处理,没有充分利用CME高阶统计信息.而独立分量分析ICA（Independent Component Analysis）,引入高阶统计量,能够分离出统计独立的非高斯信号.以南极半岛地区的15个GNSS站点为例,由于某些站点存在强烈的局部效应,因此引入了因子分析法首先对异常站进行剔除,然后对比分析了PCA和ICA方法在南极半岛地区区域滤波结果.结果显示,ICA的滤波效果要优于PCA,ICA滤波前后E、N、U三个方向RMS平均降低44.69%、26.94%、34.87%,不确定度分别降低37.43%,44.58%,55.86%,有效的降低了GNSS残差序列的发散性和速度的不确定度.

     

Spatially weighted principal component analysis (PCA) method for water quality analysis

Incorporating spatial information data into the principal component analysis is of importance. Some proposed methods of spatial weighting schemes to be applied to the ordinary PCA are reviewed and a new version of the method is proposed in the context of spatial analysis for geospatial multivariate analysis. In view of spatial variations in the hydrochemistry of rivers such combined version of the technique might be useful for reliable estimates.     

基于经验模态分解及独立成分分析的微震信号降噪方法

针对微震信号具有高噪声、突变快、随机性强等特点,基于经验模态分解(EMD)及独立成分分析(ICA)提出一种微震信号降噪方法.首先,对含噪信号进行EMD分解,获得一系列按频率从高到低的内蕴模态函数(IMF),利用原信号与各IMF之间的互相关系数辨识出噪声与信号的分界,将分界之上的高频噪声滤除;其次,为有效去除分界IMF中的模态混叠噪声,基于ICA算法对分界IMF进行盲源分离,提取其中的微震有效信号,并将其与剩余的IMF累加重构,从而得到降噪后的微震信号;最后,利用快速傅里叶变换(FFT)时频谱对比分析降噪前后的信号特征,定性说明本文方法的有效性;引入信噪比和降噪后信号占原信号的能量百分比两个参数,定量说明本文方法能充分保留微震信号的瞬态非平稳特征,降噪效果明显.     

Equivalent water height extracted from GRACE gravity field model with robust independent component analysis

The Level-2 monthly GRACE gravity field models issued by Center for Space Research (CSR), GeoForschungs Zentrum (GFZ), and Jet Propulsion Laboratory (JPL) are treated as observations used to extract the equivalent water height (EWH) with the robust independent component analysis (RICA). The smoothing radii of 300, 400, and 500 km are tested, respectively, in the Gaussian smoothing kernel function to reduce the observation Gaussianity. Three independent components are obtained by RICA in the spatial domain; the first component matches the geophysical signal, and the other two match the north-south strip and the other noises. The first mode is used to estimate EWHs of CSR, JPL, and GFZ, and compared with the classical empirical decorrelation method (EDM). The EWH STDs for 12 months in 2010 extracted by RICA and EDM show the obvious fluctuation. The results indicate that the sharp EWH changes in some areas have an important global effect, like in Amazon, Mekong, and Zambezi basins.     

基于ICA的漫射光相关谱血流分离算法研究

血流是人体的一个重要生理参数,实时测量脑部、骨骼肌及乳腺等组织的血流对疾病诊断治疗及手术、重症监护有重要意义。近红外漫射光相关谱（DCS）是新兴的组织血流测量技术,利用DCS技术进行血流测量时,每个距离的光源-探测器（S-D）均含有不同程度的表层组织和深层组织的混合信号,其中表层信号对提取深层组织的血流有较大影响。本文结合N阶线性算法（NL算法）和独立成分分析算法（ICA）对DCS技术获取的近距离和远距离光学信号进行分离处理。计算机仿真表明,本文提出的算法可以较好地分离出表层和深层组织的血流信号,对今后DCS技术在临床的血流测量应用有重要潜力。      

Classification of seismic signals at Villarrica volcano (Chile) using neural networks and genetic algorithms

Each volcano has its own unique seismic activity. The aim of this work is to construct a system able to classify seismic signals for the Villarrica volcano, one of the most active volcanoes in South America. Since seismic signals are the result of particular processes inside the volcano's structure, they can be used to forecast volcanic activity. This paper describes the different kinds of seismic signals recorded at the Villarrica volcano and their significance. Three kind of signals were considered as most representative of this volcano's activity: the long-period, the tremor, and the energetic tremor signals. A classifier is implemented to read the seismic registers at 30-second intervals, extract the most relevant features of each interval, and classify them into one of the three kinds of signals considered as most representative of this particular volcano. To do so, 1033 different kinds of 30-s signals were extracted and classified by a human expert. A feature extraction process was applied to obtain the main characteristics of each of them. This process was developed using criteria which have been shown by others to effectively classify seismic signals, based on the experience of a human expert. The classifier was implemented with a Multi-Layer Perceptron (MLP) artificial neural network whose architecture and training process were optimized by means of a genetic algorithm. This technique searched for the most adequate MLP configuration to improve the classification performance, optimizing the number of hidden neurons, the transfer functions of the neurons, and the training algorithm. The optimization process also performed a feature selection to reduce the number of signal features, optimizing the number of network inputs. The results show that the optimized classifier reaches more than 93% exactitude. identifying the signals of each kind. The amplitude of the signals is the most important feature for its classification, followed by its frequency content. The described methodology can be used to classify more seismic signals to improve the study of the activity of this volcano or to extend the study to other active volcanoes of the region.     

利用独立成分分析检测2004年和2012年印度洋地震的重力变化

2004年12月苏门答腊发生M_W9.3地震,造成巨大的质量重新分布.利用GRACE卫星月重力场数据计算了研究区域地面1°×1°网格点上的重力变化时间序列,采用主成分分析和独立成分分析两种方法,提取了重力变化的空间与时间特征,结果显示震中两侧区域的重力变化呈两极分布,其中东侧重力下降,西侧重力增加.相较于传统的主成分分析方法,独立成分分析能更好地从原始信号中提取地震的信息,能分解出具有显著阶跃变化的独立成分.除了2004年的苏门答腊大地震外,独立成分分析还分解得到了2012年Andaman地震的特征,与该次地震的空间特征与时间序列基本一致.相较于适合定量分析的多项式拟合方法,独立成分分析更适合大范围区域的定性分析,建议将两种方法相结合,取长补短,从而为GRACE地震监测提供一种更为客观、有效的方法.

     

High resolution temperature monitoring in a borehole,detection of the deterministic signals in noisy environment

Temperature was monitored as a function of time at several selected depth levels in a slim experimental borehole. The hole is 15 cm in diameter, 150 m deep, and effectively sealed from the influx of ground water by a plastic tube of 5 cm diameter. The mean temperature gradient is 19.2 mK/m. The borehole was drilled in 1993 and has been in equilibrium since then. The data obtained reveal that: (1) the temperature-time series showed a complex, apparently random oscillation pattern with amplitudes of up to 25 mK; (2) irregular temperature variations characterized by larger oscillations may alternate with relatively “quiet” intervals; and (3) the character of the oscillation may vary both in depth as well as in time and the transition between two distinct regimes may be sudden.The Fourier analysis detected “red noise” behavior of the signal but did not highlight any specific peak(s) corresponding to periodicity in the measured temperature series. We employed a variety of techniques (roughness coefficient, local growth of the second moment, recurrence and cross recurrence plots) to reveal the deterministic framework of the system behavior. All above methods were proven to be quite robust in the face of noise, and enabled the discovery of structures hidden in the signals produced by complex natural processes.Statistical analysis suggested the existence of a quasi-periodic intra-hole oscillatory convection. The temperature field in the hole has a dual-frequency structure, in which short period oscillations of about 10–30 minutes are superposed on longer variations of up to several hours. At certain conditions, so far not fully understood, the temperature oscillations may practically stop. The temperature remains within 1–2 mK for a period of several days when the oscillation pattern (convection ?) suddenly resumes.     

Detection and identification of seismic signals recorded at Krakatau volcano (Indonesia) using artificial neural networks

The Anak Krakatau volcano (Indonesia) has been monitored by a multi-parametric system since 2005. A variety of signal types can be observed in the records of the seismic stations installed on the island volcano. These include volcano-induced signals such as LP, VT, and tremor-type events as well as signals not originating from the volcano such as regional tectonic earthquakes and transient noise signals. The work presented here aims at the realization of a system that automatically detects and identifies the signals in order to estimate and monitor current activity states of the volcano. An artificial neural network approach was chosen for the identification task. A set of parameters was defined, describing waveform and spectrogram properties of events detected by an amplitude-ratio-based (STA/LTA) algorithm. The parameters are fed into a neural network which is, after a training phase, able to generalize input data and identify corresponding event types. The success of the identification depends on the network architecture and training strategy. Several tests have been performed in order to determine appropriate network layout and training for the given problem. The performance of the final system is found to be well suited to get an overview of the seismic activity recorded at the volcano. The reliability of the network classifier, as well as general drawbacks of the methods used, are discussed.     

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.     

改进的确定性地震危险性分析方法及其应用

根据改进的基于振型叠加(mode superposition)技术的地震危险性确定性分析方法,结合区域地震活动性、地质构造和地震波传播特点,通过计算理论地震图途径,分别计算了城市周边各个地震对各个场点所造成的最大影响强度(主要以峰值加速度、速度和位移体现),所确定的参数可为结构抗震设计工作提供客观依据.     

On the prediction of persistent processes using the output of deterministic models

A problem frequently met in engineering hydrology is the forecasting of hydrological variables conditional on their historical observations and the hindcasts and forecasts of a deterministic model. On the contrary, it is a common practice for climatologists to use the output of general circulation models (GCMs) for the prediction of climatic variables despite their inability to quantify the uncertainty of the predictions. Here we apply the well-established Bayesian processor of forecasts (BPF) for forecasting hydroclimatic variables using stochastic models through coupling them with GCMs. We extend the BPF to cases where long-term persistence appears, using the Hurst-Kolmogorov process (HKp, also known as fractional Gaussian noise) and we investigate its properties analytically. We apply the framework to calculate the distributions of the mean annual temperature and precipitation stochastic processes for the time period 2016–2100 in the United States of America conditional on historical observations and the respective output of GCMs.     

Sensitivity analysis of a deterministic water temperature model to forest canopy and soil temperature in Catamaran Brook (New Brunswick,Canada)

A coupled deterministic hydrological and water temperature model, CEQUEAU, was modified to include soil temperature and crown closure in its calculation of local advective terms in the heat budget. The modified model was than tested to verify its sensitivity to these modifications. An analysis of the heat budget of a small forested catchment in eastern Canada revealed that the advective term related to interflow plays a significant role in the daily water heat budget, providing on average 28% of the local advective budget (which also includes advective heat terms from surface runoff and groundwater) and nearly 14% of the total heat budget (which includes all radiative terms at the water surface, convection and evaporation, as well as the local advective terms). Relative sensitivity indices (RSIs) were used to verify the impact of the newly introduced parameters and variables. Among them, parameters related to the forest cover (crown closure and leaf area index) have a maximum RSI of ?0·6; i.e. a 100% increase in value produces a 60% decrease in the local advective term. Parameters with the greatest influence are the volume of water contributing to interflow and the amplitude of the net radiative flux at the soil surface, which, if doubled, would double the contribution of the local interflow advective term to the heat budget. Copyright © 2003 John Wiley & Sons, Ltd.     

Dyadic wavelet analysis of PDA signals

The dyadic wavelet transform is used to analyze PDA measured signals in order to identify the CASE-damping factor, which may be directly calculated from the dyadic wavelet analysis, not from the correlation study; accordingly, the pile capacity may be more exactly estimated by the CASE method. The dyadic wavelet transform can decompose a PDA measured signal into an incident impact wave and a reflected impulse wave at the certain scale that are clearly shown on the wavelet transform graph. The relation between the incidence and the reflection has been established by a transfer function based on the dyadic wavelet transform and the one-dimensional wave equation, whose phase is the time delay between the incident and the reflected and whose magnitude is a function of the CASE-damping factor. An autocorrelation function analysis method is proposed to determine the time delay and to estimate the magnitude of the transfer function that is determined by the ratio of the maximum of the autocorrelation function to the second peak value represented the reflected wave on the autocorrelation function graph. Thus, the damping factor is finally determined. An analog signal, a PIT signal and five PDA signals demonstrate the proposed methods, by which the time delay, the CASE-damping factor, and pile capacity are determined. The damping factors and pile capacity are good agreement with those by CAPWAP.     

Probabilistic risk analysis using ordered weighted averaging (OWA) operators

The concepts of system load and capacity are pivotal in risk analysis. The complexity in risk analysis increases when the input parameters are either stochastic (aleatory uncertainty) and/or missing (epistemic uncertainty). The aleatory and epistemic uncertainties related to input parameters are handled through simulation-based parametric and non-parametric probabilistic techniques. The complexities increase further when the empirical relationships are not strong enough to derive physical-based models. In this paper, ordered weighted averaging (OWA) operators are proposed to estimate the system load. The risk of failure is estimated by assuming normally distributed reliability index. The proposed methodology for risk analysis is illustrated using an example of nine-input parameters. Sensitivity analyses identified that the risk of failure is dominated by the attitude of a decision-maker to generate OWA weights, missing input parameters and system capacity.

Separation of base flow from streamflow using groundwater levels—illustrated for the Pang catchment (UK)

A new filter to separate base flow from streamflow has developed that uses observed groundwater levels. To relate the base flow to the observed groundwater levels, a non‐linear relation was used. This relation is suitable for unconfined aquifers with deep groundwater levels that do not respond to individual rainfall event. Because the filter was calibrated using total streamflow, an estimate of the direct runoff was also needed. The direct runoff was estimated from precipitation and potential evapotranspiration using a water balance model. The parameters for the base flow and direct runoff were estimated simultaneously using a Monte Carlo approach. Instead of one best solution, a range of satisfactory solutions was accepted. The filter was applied to data from two nested gauging stations in the Pang catchment (UK). Streamflow at the upstream station (Frilsham) is strongly dominated by base flow from the main aquifer, whereas at the downstream station (Pangbourne) a significant component of direct runoff also occurs. The filter appeared to provide satisfactory estimates at both stations. For Pangbourne, the rise of the base flow was strongly delayed compared with the rise of the streamflow. However, base flow exceeded streamflow on several occasions, especially during summer and autumn, which might be explained by evapotranspiration from riparian vegetation. To evaluate the results, the base flow was also estimated using three existing base‐flow separation filters: an arithmetic filter (BFI), a digital filter (Boughton) and another filter based on groundwater levels (Kliner and Knĕz̆ek). Both the BFI and Boughton filters showed a much smaller difference in base flow between the two stations. The Kliner and Knĕz̆ek filter gave consistently lower estimates of the base flow. Differences and lack of clarity in the definition of base flow complicated the comparison between the filters. An advantage of the method introduced in this paper is the clear interpretation of the separated components. A disadvantage is the high data requirement. Copyright © 2004 John Wiley & Sons, Ltd.     

Modeling spatially correlated spectral accelerations at multiple periods using principal component analysis and geostatistics

Regional seismic risk assessments and quantification of portfolio losses often require simulation of spatially distributed ground motions at multiple intensity measures. For a given earthquake, distributed ground motions are characterized by spatial correlation and correlation between different intensity measures, known as cross‐correlation. This study proposes a new spatial cross‐correlation model for within‐event spectral acceleration residuals that uses a combination of principal component analysis (PCA) and geostatistics. Records from 45 earthquakes are used to investigate earthquake‐to‐earthquake trends in application of PCA to spectral acceleration residuals. Based on the findings, PCA is used to determine coefficients that linearly transform cross‐correlated residuals to independent principal components. Nested semivariogram models are then fit to empirical semivariograms to quantify the spatial correlation of principal components. The resultant PCA spatial cross‐correlation model is shown to be accurate and computationally efficient. A step‐by‐step procedure and an example are presented to illustrate the use of the predictive model for rapid simulation of spatially cross‐correlated spectral accelerations at multiple periods.     

Principal component analysis (PCA) and mineral associations of litoraneous facies of continental shelf carbonates from northeastern Brazil

The distribution of mineral phases according to the provenance of carbonate and terrigenous facies of carbonate sediments from a large area of the continental shelf of northeast Brazil was investigated using a major element multivariate analysis approach. Heavy minerals such as ilmenite are restricted to the litoraneous facies of the continental shelf of the states of Paraíba and Pernambuco, and clay minerals are found in distal facies of the continental shelf of the State of Ceará. In the carbonate fraction, composed essentially by Mg-calcite and aragonite, there is co-variation between CaO/MgO and bathimetry in part of the studied continental shelf from depths between 15 and 20 m, apparently due to influence of the seawater temperature, degree of oxygenation and luminosity. The terrigenous facies are mainly composed of quartz, clay minerals, K-feldspars and micro-micaceous minerals, having Fe and Ti oxide and hydroxide minerals as major accessory phases. Major element behavior attests to the presence of arenaceous quartz-rich relict sediments in the 35, 60 and 80 m isobaths of the continental shelf of the state of Ceará which is here interpreted as a proxy of ancient coast lines during the Flandrian transgression.     

Loss estimation in Istanbul based on deterministic earthquake scenarios of the Marmara Sea region (Turkey)

The rapid urban development in Istanbul has lead to an increase in the exposure levels of the urban vulnerability. Due to the steadily increasing population, with improper land-use planning, inappropriate construction techniques and inadequate infrastructure systems, associated with an existing high hazard level, Istanbul is one of the most risky cities in the Mediterranean region. Estimations of casualties and losses, expected for given earthquake scenarios, are necessary to develop sustainable rehabilitation programs and for improving preparedness. Deterministic hazard scenarios and time-dependent probabilistic hazard assessment were used as input to a GIS-based loss estimation model, to evaluate the earthquake risk for Istanbul.