地统计学方法在土壤学中的应用

地统计学目前在土壤科学中得到广泛的应用和发展,成为认识土壤特征的一个重要工具.地统计学的空间变异函数和克里格插值等方法是土壤性状分析的主要手段,空间变异函数主要用于描述土壤理化性状空间变异特性,不同的插值方法可优化田间试验设计和田间采样方案,克里格插值方法则特别适用于未测量点土壤属性值的估测等.近年插值方法又被广泛应用于确定区域土壤环境容量和土壤质量标准,随机模拟则用于对土壤特性进行不确定性估计等.因而,地统计学方法对我国大量土壤学资料的整合与分析具有极大的应用前景.     

地统计学分析技术及其在气象中的适用性

首先论述了地统计学的一些特点，然后从变异函数和Kriging插值两个方面介绍了地统计学的分析技术，最后讨论了地统计学在气象研究中的适用性。     

A Combined Three-Dimensional Geological-Geostatistical-Numerical Model of Underground Excavations in Rock

Summary. This paper exploits geological and borehole geotechnical data obtained in the exploratory phase of a tunneling project to investigate in a first place if the kriging interpolation scheme may effectively reproduce the spatial variability of rock mass quality (Rock Mass Rating, RMR) in the vicinity of tunnels. For this purpose a quick solver in Fortran has been developed that performs variography analysis of 3D spatial data, fast kriging estimations of RMR between borehole sampling locations at the centroids of the elements of the numerical model, and model validation. For the purpose of an integrated underground excavation design, a step further is made by incorporating into the 3D mechanical numerical model of the rock mass, the three-dimensional (3D) solid geological model, thus coupling the geology with the ground (geotechnical) model (i.e. each element of the numerical model is assigned a geological material). The mechanical properties of each finite difference cell (or Representative Elementary Volume) of the ground model were then prescribed according to its geological type, the spatial heterogeneity of the rock mass expressed quantitatively with the kriging model, and the upscaling calculations of the mechanical properties of the intact rocks determined in the laboratory, based on the size-effect (strength dependence on size) and Damage Theory. Furthermore, a preliminary numerical simulation of the advance of unsupported tunnels in the model of the heterogeneous rock mass was performed for illustration purposes.     

A six-step practical approach to semivariogram modeling

Geostatistical prediction and simulation are being increasingly used in the earth sciences and engineering to address the imperfect knowledge of attributes that fluctuate over large areas or volumes—pollutant concentration, electromagnetic fields, porosity, thickness of a geological formation. Central to the application of such techniques is the need to know the spatial continuity, knowledge that is commonly condensed in the form of covariance or semivariogram models. Their preparation is subdivided here into the following steps: (1) Data editing, (2) Exploratory data analysis, (3) Semivariogram estimation, (4) Directional investigation, (5) Simple modeling, (6) Nested modeling. I illustrate these stages practically with a real data set from a geophysical survey from Elk County, Kansas, USA. The applicability of the approach is not limited by the physical nature of the attribute of interest.     

青藏高原气温分布的空间插值方法比较

使用反距离平方、趋势面、Kriging插值、Cokriging插值和综合方法对青藏高原1961-1990年30年平均1月气温进行空间插值比较研究，其中后两种方法能够把影响青藏高原气温分布的关键因素——高程置于插值算法之中。反距离平方和趋势面插值的结果都与实际情况相差较远；普通Kriging插值能够反映出青藏高原气温分布的一定的空间结构，但结果依然不好；由于考虑了高度变量，Cokriging插值表现出一定程度的性能改进，但因为台站海拔高度偏低，插值结果依然不理想。前4种方法效果不好的原因主要是因为青藏高原上气象台站稀少且高原西北地区无气象台站。综合方法把气温分解为结构化分量和随机分量，使用直减率把气温订正到同一海拔高度后，再对它们做Kriging插值分析，其结果较为正确地反映了青藏高原气温空间分布的特征，误差远小于其它方法。研究结果表明：样本本身的空间分布是影响插值精度的重要因素，合理的采样设计是必要的前提；对于台站稀少的地区，必须把随机插值方法和确定性方法结合起来估计气候变量的空间分布。同时给出了青藏高原1月的气温空间分布状况。     

Use of an expert system to map forest soils from a geographical information system

Abstract

Mapping forest soils using conventional methods is time consuming and expensive. An expert system is described and applied to the mapping of five forest soil-landscape units formed on a single granitoid parent material. Three thematic maps were considered important in influencing the distribution of soils. The first showed the distribution of nine classes of native eucalypt forests, and the second and third were derived from a digital elevation model and represented slope gradient and a soil wetness index combined with topographical position. These layers were input to a raster based geographical information system (GIS) and then geometrically co-registered to a regular 30 m grid. From a knowledge of soil distributions, the relationships between the soil-landscape units and the three data layers were quantified by an experienced soil scientist and used as rules in a rule based expert system. The thematic layers accessed from the GIS provided data for the expert system to infer the forest soil-landscape unit most likely to occur at any given pixel. The soil-landscape map output by the expert system compared favourably with a conventional soil-landscape map generated using interpretation of aerial photographs.     

Spatial variability in petrophysical properties in Upper Maastrichtian chalk outcrops at Stevns Klint, Denmark

The spatial variability of porosity and permeability was determined for a section in the uppermost Maastrichtian chalk exposed in the Sigerslev quarry at Stevns Klint. The aims were threefold: (1) to quantify the spatial variability pattern and its link to geology by applying systematic variogram analysis procedure, (2) to detect potential cyclicity in the petrophysical properties in this pure and overtly homogeneous chalk, and (3) to compare the section with chalk reservoirs in the North Sea in order to asses to what extent this onshore sequence has a potential as an analogue of the offshore, more deeply buried chalk reservoirs. The interval is of comparable stratigraphic age to the uppermost reservoir zones in the Tor Formation of the North Sea chalk reservoirs. The variability and spatial correlation of porosity and permeability in both horizontal and vertical directions are interpreted in a geological context and show indications of small-scale heterogeneity at 15–25 cm scale, but the clear cyclic layering described from other chalk deposits is not recognised at this locality. The investigated outcrop is not a close analogue to North Sea reservoir chalk, but some aspects are common including basic material properties, porosity/permeability trends and the variability pattern. The outcrop has a potential as analogue for some of the onshore subsurface chalk successions at shallow burial depth that form important aquifers.     

Analysis of small-scale heterogeneity in clastic rocks by using computerized X-ray tomography (CT)

The aim of this study is to recognize the basic clastic rock types on the basis of Hounsfield Units (HUs) by using statistical methods (hypothesis tests, distribution fitting, and confidence intervals). How does this recognition depend on depositional history of samples tested? Does the numerical pattern of HUs obtained from small-scale analyzing of some particular sedimentary structures coincide to the textural expectations of those sedimentary structures? Are the HU values provided by the CT measurements capable of evidencing micro-cycles belonging to a particular depositional history?For testing the numerical recognition, we analyze macroscopically homogeneous argillaceous marl, siltstone and fine sandstone samples with similar depositional background, age, and degree of diagenesis. The preliminary research shows that the Hounsfield Units can identify different sedimentary rocks considerably well. There are significant differences among the confidence intervals belonging to the different rock types. However, there also have been some slight overlaps among them.For analyzing the effects of the depositional processes in case of a particular rock type, samples from channel sandstone, distributary mouth-bar, massive (structureless) sandstone from a delta-fed turbidity fan, and channel sandstone with traction carpet origin from a delta-fed turbidity fan (all coming from Pannonian basin filling series) are used. The results prove again the significant differences in the terms of both confidence intervals and distribution-types; however, some overlapping also occurs.     

辽宁弓长岭一矿区铁矿石品位分布的地质统计学研究

弓长岭铁矿地质条件复杂,使矿山生产长期受到制约。运用地质统计学的方法对该矿区的ＴＦｅ,ＦｅＯ品位分布规律进行了系统研究,确定了品位分布的结构模型,对今后矿山开发和生产提出了建议。     

基于FFT-MA谱模拟的快速随机反演方法研究

虽然基于地质统计学的随机反演方法能够有效融合测井资料中的高频信息,但计算效率低,占用内存大,限制了它在实际资料中的应用.本文在保留传统随机反演方法优点的基础上,创造性地引入傅里叶滑动平均(Fast Fourier Transform-Moving Average, FFT-MA)谱模拟进行频率域的地质统计模拟,并利用逐步变形算法(Gradual Deformation Method,GDM)确保模拟结果与实际地震数据的匹配,构建了基于FFT-MA谱模拟的新的快速随机反演方法.与常规随机反演相比,新方法不仅分辨率高,而且能够使反演解得到快速收敛,有效提高计算效率,减少内存占用.模型试算获得了与理论模型吻合度较好的高分辨率反演结果.实际资料分析也表明新方法所得到的高分辨率反演结果能够对薄互储层进行良好的展示,为薄储层的识别提供高效可靠的技术支持.