The present work continues advanced spatial data analysis of surface contamination by radionuclides after severe nuclear accident on Chernobyl NPP. Feedforward neural networks are used for the Cs137 and Sr90 radionuclides prediction mapping and spatial estimations. Neural networks are used to model complex trends over the entire region. Residuals are analyzed with the help of geostatistical approach within the framework of NNRK (neural network residual kriging) model. Another set of data is used to validate obtained results.

View full abstract

Contour maps are available to cease overall features of physical quantity, such as elevation, temperature and so on, distributing in a specified area and based on measurement at finite points. If data for contour mapping are obtained from an area interrupted geologically by deformation, e.g. faulting and/or fracturing, it is desirable that such geological heterogeneity is taken into account when the contour map is made. Therefore, in order to make contour maps under the consideration of such geological heterogeneity, we developed a method for contour mapping by using algebraical expressions for lines of formation boundaries and complex functions for contour lines.
When estimating potential of geothermal resources, it is necessary to know distribution of initial temperature, namely rock mass temperature in a natural state just before development of geothermal resources. For this reason, we applied the method to contour mapping of the distribution of the initial temperature in the Matsukawa geothermal field and consequently obtained a reasonable result reflecting the geologic information of this field.

View full abstract

Reservoir which has been subject to brittle deformation contains fault s with displacements from a few millimeter to tens and hundreds meters, but only part of these faults may be mapped by seismic survey. Despite the complexity of the faults and fault system, two major factors, the resolution of detection tool and the grid density of line search, influence the detectability of a fault . This paper discusses the impact of seismic grid pattern on the unmapped faults and proposes the use of Horvitz-Thampson estimator for estimation of number and sizes of fault unmapped by seismic survey in a region under investigation. Computer generated fault populations in space are then used to illustrate the application of this probability model.

View full abstract

Though groundwater and rainfall correlate with each other, the recharge from rainfall into the groundwater system depends on other terrain elements. In the present study, the terrain parameters such as lineament density, drainage density, slope percentage, thickness of soil, weathered and jointed zone, average annual rainfall and mean groundwater level were subjected to multivariate analysis. The interrelations of the variables were brought out.Functional models have been developed to understand the functions of groundwater level.

View full abstract

A knowledge of statistical analysis plays an essential, role in expressing varied characteristics in quantity and revealing informations about ore deposits genesis. In this regard, we have applied some statistical methods to grade of orebodies in the Liwu copper deposit in order to establish the characteristics of grades for a metallogenetic model. Three data sets were analyzed. Data set 1 contains assays of orebodies A₁A₂B₁B₂ from the upper mineralized zone, data set 2 contains assays of orebodies B₃B₄B₅CD from the middle zone, and data set 3 contains assays of orebodies E₁E₂E₃E₄ from the lower zone.
The frequency distributions of grade data suggest that all distributions are highly skewed and theoretically close to the lognormal. This trend of grade for the deposit strikingly indicates a main metallization phase, slightly superimposed by the later metallizing process. Meanwhile, the statistical characteristics indicate clearly that the middle mineralized zone represents the most profitable section of this deposit. The statistical tests concerning grade variances and means manifest that orebodies from the three zones were formed from the same mineralization process, rather than different ones developed by some previous researchers. Consequently, three zones are different sections of a mineralized zone, which strongly agrees with a recent structural viewpoint that orebodies of the Liwu copper deposit are controlled by one homocline recumbent fold. The grade model analysis shows that there is a strong similarity between the Liwu copper deposit and Besshi massive sulfide copper deposits.

View full abstract

A study was conducted for a granite quarry producing dimensional stone in the Laurentides mountains (Québec, Canada) to forecast the recovery and waste amounts of blocks due to the presence of fractures. In this paper, the term “fracture” refers to any kind of natural structural discontinuities, i.e. fractures, joints, faults, veins, dykes. Equal area stereographic projection of fracture pole density was used to identify four sets of structural discontinuities in the quarry. For each of these sets, four binary indicator functions were defined as different levels in the quality of the rock mass. The value of these binary functions is deduced from two fracture density parameters calculated in two-dimensional unit cell: the number of fracture intersections and the cumulative fractured length. The database underlying the estimation of these parameters was obtained from a systematic field survey of fracture traces observed on the accessible working faces and floors.
A quality index was estimated on a regular grid of elementary blocks according to a systematic production plan. The predicted values of the quality index for each set were combined to estimate a recovery ratio in three target areas. It was found that the geostatistical procedure presented in this paper can be used to optimize both the extent and the orientation of the working faces. The recovery index was estimated at 15% for the first quality blocks and 15% for the second quality blocks in a 40m x 30m x 6m volume of rock. This result indicates very high ratio for the building stone industry, but the selected volume is not sufficient to ensure quarry operation for several years. Other target areas were rejected and an exploration program is currently yield on the same property, but outside the actual limits of the quarry to find additional granite reserves.

View full abstract

It has been found that after introducing two parameters, the Pareto distribution model can accurately describe the statistical characteristics of the observed pool size distributions with progressive increases in exploration maturity. Based on several exploration exampies one may analyze the trend of the distribution parameters with changes in exploration maturity, thus estimate the parent pool size distribution. The proposed generalized Pareto distribution model is a dynamic method that establishes the relationship among the pool size distribution, exploration maturity (human factor), and basin richness (geological factor) ) . This model can be used to estimate pool size, number of pool and to predict the amount of exploration effort of finding those pools in future exploration. The method has been applied to quantitative resource assessments in Songliao basin and Sichuan basin.

View full abstract

Multi-purpose hazard map for slope failures with various information which can be available for many people has been studied from data treatment. To make these information maps covering various time range, a processing system for data transformation is required. Probability of slope failures is generally believed to depend on angles of slopes. However, it is generally considered to be changeable in terms of time, and its changing style may also depend on rock types. Under some assumption, we can estimate the probabilities at any spot x, y, and express them as the function D (x, y, t) .
As a case study, an attempt has been made to draw hazard map of the Kagoshima area on the basis of topographical and geological data both of which were digitized. Several maps drawn through the system enable us to understand dangerous spots and their changes in terms of time graphically. Because slope angles and rock types distributed there are different place to place and probabilities also depend on them, their changes seems to be complicated. Consequently, these maps and the system may be available for various purposes of users.

View full abstract

Methodology involving four steps for data base generation required for LHZ map of Western Ghats has been suggested. The first three steps viz. i) analysis of previous data, ii) laboratory investigations for preparation of base maps and iii) field checks offer valuable information on causative factors and their components of slope instability. This information, in turn, assisted in assigning the rating to each component or group of components in the fourth step. The rating scheme helped in obtaining the values of total hazard evaluation factor, on the basis of which the study area has been categorized into four landslide capability zones namely, Most Hazardous-, Hazardous-, Moderately Hazardous- and Least Hazardous zones. The numerical data obtained in a representative area namely the Kera area has been presented and efficacy of the methodology is discussed. It has been suggested that the rating scheme can be useful in correlation of landslide affected areas of Western Ghats and the data base can be used in GIS/ILWIS.

View full abstract

It is important to predict the occurrence of debris flow for engineers and engineering geologists. Especially the prediction of occurrence point of debris flow involves geological and geomorphological problems. When geomorphological development is controlled by geological feature, it must be possible that geomorphological analysis predicts the point where debris flow will occur. In this study, the analysis of geomorphology using fractal is attempted as fundamental work for the prediction of place where debris flow will occur. When the analysis is compared with topography and geology, it is shown that the fractal geometry is dependent upon the topographical feature and the geological structure at mountains regions. In these regions the debris flows occurred at the place where the contour of dimensions is dense. Therefore, if DEM (Digital Elevation Model) data will be made in more detail, it should be possible to predict the debris flow occurrence point.

View full abstract

In this study the two dimensional hydrodynamic model to simulate the propagation of flood shock on the steep mountainous area have been developed. This model is formulated by a hydrodynamic system with the assumption of unsteady flow as steady flow which can reduce computation time. The numerical computation for this model is executed by use of the eight point implicit scheme. The Incremental Dynamic Programming (IDP) technique is adopted for the optimization of grid size during the computation. The model is applied to compute the propagation of flood shock on the steep mountainous area located in southern part of the Korea. Comparison of the results between the proposed assumption model and the general type of two dimensional model indicates the proposed model is nearly as good as, and computationally much faster, than the general two dimensional.

View full abstract

Calculation of the occurrence criteria and flow rate of debris flow on a slope may be possible if the values of the hydraulic conductivity, the depth of the deposited material, the length and the gradient of the slope are given. However, the obtained values on the slope are generally not accurate enough for practical use due to the error in the measurements. On an active volcano debris flow, these problems may be more severe. In this study, neural networks are introduced into prediction of occurrence and amount of the debris flow. Neural networks are expected to learn general rules of the debris flow without the hydraulic, hydrologic and geologic factors, and to predict the occurrence and amount of the debris flow objectively from rainfall data. The proposed methods are checked by applying to the debris flow in Unzen volcanic area in Japan. The results of the application of the predicting model lead to the conclusion that the neural network models are useful tool for prediction of debris flow.

View full abstract

A mathematical model is proposed to determine hazard zone of debris flow and debris deposition. Two layer high concentrated model of solid transport is used in cooperation with continuity equation of solid phase for determination of deposition in the model. Transition from supercritical flow to subcritical flow exist from mountainous stream to flood plain. The proposed model simulated well such transition. Here, the simulated deposition bed are compared with experimental ones to verify the model. Simulated results show fairly good agreement with experimental results. The flow properties and deposition of debris flow in the flood plain can also be calculated by suggested model. The optimum value of the parameter α (the parameter of the solid transport equation) is found to be around 0.13 for determination of deposition of debris flow. Simulated deposition and debris flow surface cap be used to determine hazard zone and simulated flow velocity may be applied to decide hazard risks of different areas. In this respect, the hazard zone of debris flow and its deposition can be determined in the different flow conditions of debris flows.

View full abstract

The treatment of the obtained data which is easy to be collected from the weathering crust in north-west region of North Vietnam and the UTM maps by statistical analysis could contribute to the quantitative characterization of the area with high potential for landslides related to the weathering crust. The functions, established on the basis of statistical analysis of the natural factors affecting landslides related to the weathering crust, could be used for identifying the area with high potential for landslide, forecasting the occurrence of landslide, occurrence of each type of landslides related to the tropical weathering crust in the region with similar natural conditions as north-west Vietnam.

View full abstract

Nearly all engineering structures are supported in some way on natural earth materials; this inevitably creates an interaction between geological conditions and foundation analysis, design and construction. Thus, both geologists and geotechnical engineers have a vital role to play in many engineering operations which interact with the ground. In this paper, the interplay between geotechnical engineers and geologists is discussed.
To analyse and design geotechnical engineering works in urban areas, the practice of geotechnical engineers includes the study of the mechanical properties of natural earth materials and, very often, the use of mathematical models to describe the fundamental behaviour of these materials. With the help of recent advances in techniques of site investigation and laboratory testing, new insights into soil stress-strain relationships at small strains and recent stress history effects have been gained and are introduced in this paper. In addition, some mathematical soil models such as the non-linear Brick model, which has the capability to capture these new observed soil behaviour, is discussed and examined. The use of geological information in assisting the back-analysis of a case history in the city centre of Cambridge is illustrated. The ongoing processes of field monitoring and back-analysis can help to refine and improve our understanding of the behaviour of soil structure interactions, which in turn provides guidance for future design. It is intended that through the introduction, discussion and illustration, geologists will have a better understanding of soil behaviour. It is hoped that this understanding will help to create a better partnership between geologists and geotechnical engineers in the future.

View full abstract

In a study of reservoir heterogeneity within a Lower Mississippian sandstone in central West Virginia (United States), we examined an oil field that occurs within a plunging syncline and shows evidence of structural heterogeneity resulting from faulting, and lithologic heterogeneity resulting from deposition in a fluvial-deltaic environment, with subsequent diagenesis. Visualization of the latter type of heterogeneity took the approach of three-dimensional modeling of porosity and electrofacies using kriging.
Some kind of reconstruction of original bedding relationships needed to be made prior to variography. The reservoir sandstone lies just below an angular unconformity; many wells were not drilled far enough below the sandstone to penetrate a deeper unit that could have acted as a datum. Using a published algorithm for stratigraphic correlation and digitized gamma-ray logs, the relative vertical shift between pairs of wells giving the highest crosscorrelation was computed and stored in a square matrix. Multidimensional scaling of the matrix yielded a vector of values representing the vertical shift necessary to adjust the relative elevation of each well with respect to sea level. Density logs were also digitized and used to estimate porosity. Cluster analysis of depth-specific gamma-ray and density log responses from cored wells showed the presence of at least four electrofacies that could be matched through core descriptions with environments of deposition. Discriminant functions calculated for these groups were used to classify log responses from uncored wells. Kriging and conditional simulation of log porosity and electrofacies followed conventional variography. Slice maps and cross sections clearly showed the complex facies relationships, and large variation in porosity between closely-spaced wells.

View full abstract

This paper shows the idea and FORTRAN-77 source code of the division of the digital grid data set into an aggregate of triangular elements to calculate a gravity anomaly. Gravity effect calculated by a sphere model using the technique in this exhibited study was corresponding to the analytic solution well. Because of having come to be able to calculate appropriately and efficiently, calculating the repetition of the correction of the model became comparatively easy. The example that is computing results of the basement structure in Osaka Plain, South-west Japan, by the repetition is shown in this paper.

View full abstract

In this paper, the Bazhitou-Chaye Permian Section of Guangshun County in Guizhou Province of China is chosen as the representative section, in which five chemical stratigraphic units and seven abnormal strata were divided by the multivariate statistical analysis using 17 trace elements, As, B, Ba, Co, Cr, Cu, Ga, Mn, Mo, Ni, Pb, Sn, Sr, Ti, V, Zn and Zr of 59 layers as the variables.
Through the field observation and sedimentary lithofacies analysis, two bigger sedimentary cycles and nine second-order cycles of the Bazhitou-chaye Permian Section were recognized in the light of diagenesis, color, microf acies, thickness of the layers as well as their stacking patterns.
According to the comparison between chemical stratigraphy and sequence stratigraphy, the geochemical response to the Permian sea-level change and the sedimentational environments in the Upper Yangtze Platform are discussed.

View full abstract

The Middle Siwalik molasse sequence of the study area comprises an interbedded assemblage of coarse to fine sandstone, siltstone, arenaceous shale and mudstone of variegated colors. A statistical approach using quasi-independence and entropy models was used to show that the Middle Siwalik facies developed distinct cyclic order during deposition. The upward transition path of lithologic states typical for this sequence is as follows: coarse sandstone (COSD), medium to fine sandstone (FMSD), interbedded sandstone/siltstone (SLSD), mudstone (MDST) and again coarse sandstone (COSD) . These recurring fining upward asymmetrical cycles correspond to the development of basal channel bars, succeeded by levees, than by overbank vertical accretion mudstone which are often topped by calcrete.
Entropy plots of the E ^(pre) and E ^(post) values of each lithological state, corresponds to the Type A-4 category of Hattori (1976) which signifies truncated asymmetrical cycles. The value of E (system) falls within the fluvial-alluvial envelope of Hattori and consistent that the Middle Siwalik molasse conceived in a fluvial environment. It is suggested that autocyclic switching of braided river channels in response to gradual subsidence on the flood plain is the most likely cause for the development of asymmetrical cycles in this fluvial depositional model.

View full abstract

Ore segregations of chromian spinels are located mainly in layered ultramaficmafic complexes and in ophiolite sequences. Numerous observations on compositional variability of chromites have been made, and various aspects of mineralizations such as genesis, stratigraphic position, tectonic setting, texture, and host rock are reflected in the major element chemistry of individual ore bodies. However, conventional geochemical classification schemes based on only a few components prove to be uncertain because of overlaps between groups.
Published microprobe data on the composition of more than 400 ore and accessory chromite grains from mineralized complexes of most of the important chrome-producing countries are evaluated by means of multivariate mathematical methods in order to reveal the relationships between mineral chemistry and other features of chromite deposits. Cluster analysis was used to separate, and non-linear mapping and principal components analysis to visualize natural geochemical classes. Multiple discriminant analyses based on different sets of groups led to plots enabling effective classifications of chromites.

View full abstract

Importance of pressure transient analysis during preparation of development schemes and performance analysis of oil/gas reservoirs is accredited in oil and gas industry due to its usage in estimation of petrophysical parameters of the reservoir and analysis of inflow performance relationship of oil wells, open flow performance of gas wells etc.
In general, data analysis may be stated as a two-fold process: (i) data management, selection and calculations and (ii) plotting of the required data. Most of the available PC based software packages for pressure transient analysis perform complex calculations but are deficient in their data management / graphic output capabilities which usually is rigid and limited. Based on the popular spreadsheet software package Quattro Pro^TM, a powerful and timesaving spreadsheet program for pressure transient analysis is presented. The spreadsheet prepared is simple to use and can be customised to user's needs by modifying the given format with a provision of further expansion.

View full abstract

The Bihar Mica Belt (BMB) -largest repository of muscovite in India, is located across the central part of Bihar State. Extending ENE-WSW for over 150 Km and having 20-40 Km width, BMB marks the eastern extremity of the Proterozoic Satpura Orogenic Belt of central India. Three groups of pegmatites recognised from this belt are: Gr. A or deformed pegmatite-characterized by deformed mineral grains-often foliated and are barren of commercial muscovite; Gr. B or muscovite pegmatite-typically zoned with quartz core and are the chief source of muscovite; Gr. C or rare metal pegmatite-free of deformation, characterized by dominance of K-feldspar over plagioclase, frequent presence of cleavelandite, lepidolite and muscovite.
Multigroup Discriminant Analysis using 6-element (viz. Na, K, Li, Rb, Ba and Pb; all in ppm) contents of K-feldspar from Gr.A, B and C pegmatites has been carried out. The two most influential discriminant axes (F₁ and F₂) together account for 99.7% of the total dispersion of the pooled data matrix. On the F₁-F₂ plane the projection of the samples forms separate clusters with little overlap. Furthermore the F₁ axis alone accounts for c. 94% of the total dispersion and using only this function the three groups of pegmatites could be discrimmated.

View full abstract