近红外反射光谱学在岩石矿物研究中的应用Ⅲ．快速测定地质样品中二氧化碳

使用一组不同岩石矿物性质的25个地质样品,研究了地质样品中二氧化碳含量和近红外漫反射光谱的关系。利用多重散射光校正、标准正常变量转换及导数光谱等化学计量统计学方法,预处理了漫反射光谱数据,获得了偏最小二乘法(PLS)回归的多变量校正方法,建立了应用近红外(NIR)反射光谱学方法无损快速测定各种地质样品中二氧化碳模型。从光谱特性方面解释了大多数进入多变量校正模型的波长与碳酸盐矿物的CO2-3基团光谱吸收有关。NIR估算结果表明,应用二氧化碳校正模型来定量分析地质样品中二氧化碳的方法是可行的。     

Pollen percentages,tree abundances and the Fagerlind effect

Pollen diagrams traditionally are read as if pollen percentages were linearly related to relative tree abundances, although the slopes and intercepts of these relationships are accepted to differ among taxa. Corresponding map patterns of modern pollen and tree percentages support this assumption of linearity, which also underlies the use of linear regression on percentage data for pollen-tree calibration. Fagerlind, however, showed that the theoretical relationship need not be linear and may be confounded by interdependencies among taxa. Regressions and scatter plots of modern pollen and tree percentages are here compared with results of extended R-value (ERV) models, which correct for the ‘Fagerlind effect’. Three data sets from Wisconsin and Michigan, USA illustrate that regression coefficients provide a first approximation to their ERV equivalents, but scatter plots derived from the ERV analyses show reduced scatter about linearised relationships.     

Multivariate forecast of winter monsoon rainfall in India using SST anomaly as a predictor: Neurocomputing and statistical approaches

In this article, the complexities in the relationship between rainfall and sea surface temperature (SST) anomalies during the winter monsoon over India were evaluated statistically using scatter plot matrices and autocorrelation functions. Linear, as well as polynomial trend equations were obtained, and it was observed that the coefficient of determination for the linear trend was very low and it remained low even when polynomial trend of degree six was used. An exponential regression equation and an artificial neural network with extensive variable selection were generated to forecast the average winter monsoon rainfall of a given year using the rainfall amounts and the SST anomalies in the winter monsoon months of the previous year as predictors. The regression coefficients for the multiple exponential regression equation were generated using Levenberg-Marquardt algorithm. The artificial neural network was generated in the form of a multilayer perceptron with sigmoid non-linearity and genetic-algorithm based variable selection. Both of the predictive models were judged statistically using the Willmott's index, percentage error of prediction, and prediction yields. The statistical assessment revealed the potential of artificial neural network over exponential regression.     

天然气水合物地震空白带现象正演模型研究

空白带现象是人工地震方法识别含天然气水合物沉积地层的重要标志之一。针对水合物沉积物的 3种微观模式 ,首先采用度量振幅空白程度的振幅比方法 ,计算并分析了水合物饱和度不随孔隙度变化 ,以及水合物的饱和度随孔隙度变化 (即CGHC和VGHC)两种情况 ,然后对上述两种情况设计了地震模型 ,合成了水合物沉积物地震纪录 ,获得了地震空白带现象。研究结果表明 :水合物沉积物饱和度的变化是振幅空白的主要原因。对孔隙度和饱和度都不固定的复杂情况 ,沉积物的岩性存在多解性 ,需用各种方法的多元信息进行对比研究 ,才能更好地解释空白带的产生及其特征。     

基于数字地球平台的地质钻孔三维模拟与可视化

研究了地质钻孔信息的管理和可视化技术,提出了一种在数字地球平台中进行城市钻孔信息模拟和可视化的方法。该方法基于一个统一的钻孔数据库标准,首先将钻探点位信息和地层分层信息组织成具有不同细节层次的钻探点位模型、钻孔分层散点模型和钻孔管段模型,然后构建基于细节层次模型的海量钻孔信息多尺度表达机制,最后将这些模型加载到数字地球平台中进行可视化及查询、分析。在上海市的应用实例显示,该方法实现过程简单、自动性高,生成的模型易于在国际互联网上进行分发、集成与共享。该方法的应用将有助于地质工作者向其他科技人员以及社会公众展示、分发自己拥有的钻孔信息,为以钻孔数据为代表的城市地质信息的社会化服务奠定基础。     

汉江河谷旬阳段区域滑坡规律及斜坡不稳定性预测

基于斜坡工程地质测绘和实验室分析,文中讨论了研究地段区域滑坡规律。引入信息量及二态变量回归分析建立了斜坡不稳定性两种预测模型。文章重点论证和提出了影响因素筛选、多边形网格单元划分的原则及预测精度评价的原理方法。编制了斜坡不稳定性分区图,检验表明其预测精度高;两预测分区图对比表明有很好的一致性。     

Cumulative semivariogram models of regionalized variables

The cumulative semivariogram approach is proposed for modeling regionalized variables in the geological sciences. This semivariogram is defined as the successive summation of half-squared differences which are ranked according to the ascending order of distances extracted from all possible pairs of sample locations within a region. This procedure is useful especially when sampling points are irregularly distributed within the study area. Cumulative semivariograms possess all of the objective properties of classical semivariograms. Classical semivariogram models are evaluated on the basis of the cumulative semivariogram methodology. Model parameter estimation procedures are simplified with the use of arithmetic, semilogarithmic, or double-logarithmic papers. Plots of cumulative semivariogram values vs. corresponding distances may scatter along a straight line on one of these papers, which facilitates model identification as well as parameter estimation. Straight lines are fitted to the cumulative semivariogram scatter diagram by classical linear regression analysis. Finally, applications of the methodology are presented for some groundwater data recorded in the sedimentary basins of the Kingdom of Saudi Arabia.     

不同方法在感潮河段ADCP在线测流系统中应用的比较分析

在水力因素多变的长江下游感潮河段建立ADCP在线测流系统实时采集指标流速,可选用合适的方法推求断面平均流速,从而实现流量的实时报汛和整编。多元线性回归分析和BP神经网络具有原理明晰、实现便捷等特点,为比较以上两种方法在断面平均流速计算中的优劣,以南京水文实验站2014年以来实测数据为例,分析不同情况下两种模型的拟合精度和预测精度。结果表明,两种模型均具有较好的有效性、精确性和稳定性,且拟合精度与模型选用的监测指标有关;对于只采用单一指标流速而言,BP神经网络模型的结果明显优于多元线性回归模型。同时,两种模型都能较好的预测断面平均流速,其中BP神经网络适应更好。     

近红外漫反射光谱在沉积物化学成分分析中的研究进展

沉积物中有机质及相关元素含量、重金属含量等是研究环境污染和古环境的依据。利用近红外漫反射光谱测定沉积物中的化学成分,方法简便、快捷、价廉。本文概述了近红外漫反射光谱测定沉积物中化学成分的基本流程、样品制备及建模方法,介绍了如何通过选择建模样品、利用光谱预处理及回归分析等方法提高近红外光谱模型的定量能力,综述了近红外漫反射光谱分析沉积物中的有机碳、总氮、总磷、生物硅、重金属含量等方法。但是利用近红外光谱分析沉积物中的化学成分研究起步较晚,仍存在很多问题,有必要深入探讨近红外光谱分析沉积物的化学成分时产生误差的机理,进一步开展建模样品的计算机选择方法、光谱预处理方法和数学模型优化等方面的研究,提高近红外漫反射光谱分析沉积物中化学成分的精密度和准确度。     

基于广义回归神经网络的砂土液化综合判别方法

基于厦门地区大量钻孔试验数据,分别采用规范法和Seed法对该区饱和砂土进行液化判别。然后选取二者判别结果相同的数据作为训练和测试样本,运用广义回归神经网络,对二者判别结果分歧的钻孔数据进行二次判别。结果表明:广义回归神经网络性能良好,预测准确度高。此外,这种综合判别方法也提高了饱和砂土液化判别的准确度,并为其他地区饱和砂土的液化判别研究提供借鉴和参考。      

边坡可靠度分析的一种新的优化求解方法

介绍了Low & Tang提出的一种新的可靠度优化求解方法,并将之用于边坡可靠度分析中：该方法适用于任何概率分布的相关变量,不必计算当量正态均值和方差、相关变量独立变换,直接在变量的原始空间内搜索边坡的最小可靠指标和概率临界滑面,可采用任何合适的约束优化方法进行求解,方法清晰简洁。边坡可靠度分析常用的滑面有2个：最小安全系数（变量均值处）对应的确定性临界滑面和最小可靠指标对应的概率临界滑面,但这2个滑面在有些情况下差别较大,Hassan & Wolff提出了一种简化方法可以方便地获得概率临界滑面,但由于方法简单,受到质疑。通过一系列算例分析,优化求解方法得到的概率临界滑面和Hassan & Wolff的简化方法滑面非常接近,显示了简化方法的有效性,值得在工程实践中推广。     

基于投影寻踪回归的指标规范值的水质评价模型

传统的投影寻踪回归(PPR)的矩阵表示法用于水质评价,当指标较多时,不仅优化参数矩阵元的学习效率低,而且优化效果亦受到影响。若适当设置3类水体(地表水、地下水和富营养化水体)各指标的参照值及指标值的规范变换式,使不同指标的同级标准的规范值差异不大,从而可以认为用规范值表示的不同指标皆与某个规范指标"等效"。因此,只需构造并优化得出对各指标规范值都共同适用的2个指标变量的NV-PPR(2)和3个指标变量的NV-PPR(3)模型,对于指标变量较多的NV-PPR建模,只需将其分解为若干个NV-PPR(2)和(或)NV-PPR(3)的组合表示即可。对模型的实用性进行的效果检验表明:基于指标规范变换的3类水体的水质评价的投影寻踪回归模型具有形式简洁、计算简便和普适通用的特点。     

缝洞型油藏连通单元地质建模研究——以哈拉哈塘油田为例

针对哈拉哈塘油田奥陶系裂缝孔洞型储层连通单元井组注水效果差异性很大的问题,开展缝洞单元连通井组地质建模研究。哈拉哈塘油田储集空间岩溶洞穴、孔洞、裂缝非常发育,类型形态多变,具有极强的非均质性,储层空间缝洞体识别及连通方式预测与定量评价难度大。本文首先利用地震识别的大型溶洞和蚂蚁体追踪的大尺度裂缝分布,通过确定性建模方法,建立离散大型溶洞模型和离散大尺度裂缝模型;然后在储层的波阻抗数据体和缝洞体储层构造模型的空间约束下,建立缝洞体连通单元的储层孔隙度模型,储层渗透率模型主要通过线性回归建立基质孔隙度、裂缝参数与渗透率关系,结合动态数值试井对静态回归计算的渗透率数值进行修正;最后利用曲率属性形象表征微裂缝,采用多属性协同模拟方法,建立多尺度离散缝洞储集体三维地质属性模型。该方法定量刻画了缝洞储集体在三维空间的展布特征,很好地表征了单元连通井组连通方式,为单元连通井组开发奠定了坚实的地质基础。     

石油与天然气定量评价的体积统计法（续）

多元线性回归方法的可靠性如何呢?上述研究者以美国的二叠系盆地为评价对象,对该方法进行验证,所取得的结果是令人满意的,和其它方法所得的结果很接近。     

Assessing impact of urban impervious surface on watershed hydrology using distributed object-oriented simulation and spatial regression

In this study, we investigated the relationship between watershed characteristics and hydrology using high spatial resolution impervious surface area (ISA), hydrologic simulations and spatial regression. We selected 20 watersheds at HUC 12 level with different degrees of urbanization and performed hydrologic simulation using a distributed object-oriented rainfall and runoff simulation model. We extracted the discharge per area and ratio of runoff to base flow from simulation results and used them as indicators of hydrology pattern. We derived percentage of ISA, distance from ISA to streams, and stream density as the watershed characteristics to evaluate the relationship with hydrology pattern in watersheds using ordinary least square, spatial error and spatial lag regression models. The comparison indicates that spatial lag regression model can achieve better performance for the evaluation of relationship between ratio of runoff to base flow and watershed characteristics, and that three models provide similar performance for the evaluation of relationship between discharge per area and watershed characteristics. The results from regression analyses demonstrate that ISA plays an important role in watershed hydrology. Ignorance of spatial dependence in analyses will likely cause inaccurate evaluation for relationship between ISA and watershed hydrology. The hydrologic model, regression methods and relationships between watershed characteristics and hydrology pattern provide important tools and information for decision makers to evaluate the effect of different scenarios in land management.     

A Mathematical View of Weights-of-Evidence,Conditional Independence,and Logistic Regression in Terms of Markov Random Fields

New light is shed on mathematical methods of potential modeling from the point of view of Markov random fields. In particular, weights-of-evidence and logistic regression models are discussed in terms of graphical models possessing Markov properties, where the notion of conditional independence is essential, and will be related to log-linear models. While weights-of-evidence with respect to indicator predictor variables and logistic regression with unrestricted predictor variables model conditional probabilities of an indicator random target variable, the subject of log-linear models is the joint probability of random variables. The relationship to log-linear models leads to a likelihood ratio test of conditional independence, rendering an omnibus test of conditional independence restricted by a normality assumption obsolete. Moreover, it reveals a hierarchy of methods comprising weights-of-evidence, logistic regression without interaction terms, and logistic regression including interaction terms, where each former method is a special case of the consecutive latter method. The assumptions of conditional independence of all predictor variables given the target variable lead to logistic regression without interaction terms. Violations of conditional independence are compensated exactly by corresponding interaction terms, no cumbersome approximate corrections are needed. Thus, including interaction terms into logistic regression models is an appropriate means to account for lacking conditional independence. Logistic regression exempts from the burden to worry about lack of conditional independence. Eventually, the relationship to log-linear models renders logistic regression with indicator predictor variables optimum for discrete predictor variables. Weights-of-evidence applies for indicator predictor variables only, logistic regression applies without restrictions of the type of predictor variables and approximates the proper distribution in the general case.     

基于改进型拉开档次法的泥石流危险度评价实例

泥石流危险度评价是泥石流研究的热点和难点问题。运用极大不相关法剔除泥石流危险度评价中提供信息重叠过多的因子,并应用主观赋权的序关系分析法（G1法）对客观赋权的拉开档次法进行改进,从而获得了一种新的泥石流危险度评价方法。以吉林省和龙市地质灾害调查与区划中的10条泥石流危险度评价实例进行验证并与传统方法对比,结果表明：该方法操作简便,其中4条处于分级点附近的泥石流危险度比传统方法高出一个等级,改进型拉开档次法能体现出传统评价方法中同级别泥石流危险度的差异性,评价结果与实际情况吻合得更好。结论虽然偏于安全,但能为有针对地开展泥石流灾害防治工作提供更科学的指导,可为泥石流危险度评价提供一种新的评价方法。     

Mapping landslide susceptibility with logistic regression, multiple adaptive regression splines, classification and regression trees, and maximum entropy methods: a comparative study

Four statistical techniques for modelling landslide susceptibility were compared: multiple logistic regression (MLR), multivariate adaptive regression splines (MARS), classification and regression trees (CART), and maximum entropy (MAXENT). According to the literature, MARS and MAXENT have never been used in landslide susceptibility modelling, and CART has been used only twice. Twenty independent variables were used as predictors, including lithology as a categorical variable. Two sets of random samples were used, for a total of 90 model replicates (with and without lithology, and with different proportions of positive and negative data). The model performance was evaluated using the area under the receiver operating characteristic curve (AUC) statistic. The main results are (a) the inclusion of lithology improves the model performance; (b) the best AUC values for single models are MLR (0.76), MARS (0.76), CART (0.77), and MAXENT (0.78); (c) a smaller amount of negative data provides better results; (d) the models with the highest prediction capability are obtained with MAXENT and CART; and (e) the combination of different models is a way to evaluate the model reliability. We further discuss some key issues in landslide modelling, including the influence of the various methods that we used, the sample size, and the random replicate procedures.     

一种新的储量估算法—杨赤中滤波与推估法及其应用

杨赤中滤波与推估法是一种对矿床地质变量统计的新方法。为使这一新理论用于勘探网规则或不规则情况下及在矿体小而勘探抽样点比较少,产状复杂的铀矿体的储量估算,我们用此法对2个矿床的5个矿体进行了储量估算,取得了较满意的效果。实践证明,杨赤中滤波与推估法与传统的块段法比较,具有公式简明、使用方便、易掌握、计算速度快、结果准确,花费少、经济效益好等优点。本文主要介绍了此种方法实际应用的步骤,应用的体会以及效果的初步评价。     

Exploring the potential of location-based social networks data as proxy variables in collective human mobility prediction models

The study of human mobility has gained much attention in recent years. To date, various models have been developed to predict human mobility patterns for intra- and/or inter-city cases. These models incorporate the populations as proxy variables in the place of real variables which cannot be observed easily. However, inaccuracies in predicting human mobility within cities are usually encountered. One source of inaccuracies in intra-city scenarios arises from the fact that cities’ populations are influenced by people from other areas. Therefore, population cannot be regarded as a good proxy variable for movement modeling. The objectives of this article are to introduce new proxy variables for use in current models for predicting human mobility patterns within cities, and to evaluate the accuracy of the predictions. In this study, we have introduced new proxy variables, namely, venues and check-ins, extracted from location-based social networks (LBSNs). In order to evaluate the models, we have compared our results with empirical data obtained from taxi vehicles, based on trip distances and destination population distributions. The Sørensen similarity index (SSI) and R-squared measures were also used to compare the performances of models using each variable. The results show that all models with LBSN variables can capture real human movements better within Manhattan, New York City. Our analytical results indicated that the predicted trips using LBSN data are more similar to the real trips, on average, by about 20% based on the SSI. Moreover, the R-squared measures obtained from regression analyses were enhanced significantly.