جستجوی مقالات مرتبط با کلیدواژه "روش های آماری" در نشریات گروه "زمین شناسی"

In current research, primary classification of Khomein-Robat-Arregijeh Pb-Zn deposit has been ‎carried out first, using support vector machine (SVM) method. To achieve this goal, 548 ‎available multivariate data point comprising 337 induced polarization (IP) and resistivity (Rs) ‎geophysical data, 211 rock type and Pb-Zn assay data from outcrops, trenches and test pits, ‎three variables IP, Rs and rock type were considered as predictive variables as well as Pb-Zn ‎assay as target variable. Afterward through designing and training an SVM model, on the basis ‎of three cut off grade 1.5, 2 and 3 percent, the deposit was classified to two classes: high-grade ‎zone or anomaly (above the cut off grade) and low-grade zone or background (lower than cut ‎off grade). Then exploratory boreholes scattered in the region, were drilled in the locations ‎known as high-grade zone as well as anomalies of the geophysical pseudo-sections. In the next, ‎using all exploration information, designing the detailed exploration grid of the deposit was ‎performed through statistical and geostatistical methods. A rectangular grid with dimensions of ‎‎36*35m was obtained by analytical method based on classical statistics. Employing geostatistical ‎method and 3-D variography for Pb-Zn assays of the boreholes using SGEMS software ‎concluded a square grid with dimensions of 55*55m. According to the results of the research, ‎semi-deep exploration activities are suggested at the nodes of the new exploration grid, designed ‎by geostatistical method due to more accuracy of the method.‎

Tekieh Lead-Zinc ore deposit that is located in the Sanandaj-Sirjan structural zone has been recognized as one of the most important mineralized regions in Malayer-Isfahan metallogenic sub-state, south east of Arak (Momenzadeh and Ziseman, 1981). Carbonate host units have been developed along (or across) the Vishan-Tekieh anticline as the main structure extended in NW-SE trends (Annells et al, 1985). According to geochemical investigations (Salehi, 2004), the element content of the mineralized regions has originated from Alpine post-volcanisms and subsequently it has migrated toward early Cretaceous formations (dolomitic limestones) among several hypogenic stages (Torkashvand et al., 2009). Also echelon type structures consisting of folded systems and inversed faulting of structures are the most common features in western and eastern parts of ore deposit regions (Annells et al, 1985). Syngenetic enrichments beside limited (rarely developed) epigenetic mineralization have been known as two main phases which are closely relevant to ore forming processes in the massive lenses and vein type occurrences, respectively (Momenzadeh and Ziseman, 1981).
Material and In this research, two statistical techniques that consist of classical and fractal equations (Mandelbrot, 2005) were applied in geochemical (Torkashvand et al., 2009) and geophysical (Jafari, 2007) databases for obtaining the linear and nonlinear distributions of geochemical elements (Tekieh Pb-Zn content) in association with resistivity variations and induction polarization measurements (Calagari, 2010). According to linear statistical techniques (Torkashvand et al., 2009), the main central parameters such as mean, median and mode in addition to variances and standard deviations as distribution tendencies could be used for obtaining the regression coefficients of the databases. However, in fractal statistics, a reliable regression between geoelectrical - geochemical anomalies should be calculated based on measuring the fractal dimensional variations in the recursive patterns (Mehrnia, 2013). In practice, the Area-Concentration equations (Mandelbrot, 2005) were applied in resistivity, induction polarization, Pb and Zn datasets for achieving the nonlinear relationships in anomalous regions which were characterized by increasing in regression coefficients with more spatial correlation of the variable than linear statistics (Mehrnia, 2013).
This research showed that both linear and nonlinear statistics are able to estimate the spatial association of geochemical anomalies with geophysical variables. A meaningful increase in the regression coefficient was also revealed after measuring the self-similar peculiarities of concentration values on gridded plots (Salehi, 2004; Torkashvand et al., 2009). From the fractal point of view, Pb ore-minerals have been deposited in the western sub-region, while Zn mineralization seems to be extended in the depth of eastern alterations. Also a predictable geochemical zonation can be considered in the western target (meaningful Pb anomalies) that is more patterned than the eastern halos according to geological observations (Momenzadeh and Ziseman, 1981) and mineralogical evidences (Salehi, 2004). An increase in Supra ore/Sub ore proportional content was measured in the western sub-region which indicated more reliable potential of Pb mineralization (Galena as a particular indication of sulfide-rich minerals) than the same phases of ore forming processes in the eastern sub-region, although the content of Pb-ores rapidly decreases in the eastern target and is replaced by Zn minerals (Sphalerite as particular indication of sulfide-rich mineralization). Because power law relationships are significant in both geochemical and geophysical anomalies (Mehrnia, 2013) a detailed program including borehole geophysics and litho-geochemical land-surveys should be considered in the prospected regions. Therefore, upcoming phases should emphasize on self-organized distribution of Pb-Zn anomalies to introduce a new set of nonlinear distributions in order to find the confidence regression coefficients between the variables. As the final results, fractal analysis of available databases represented new target areas with better mineralization aggregations than linear analysis of the anomalous regions according to micrographs. It means that surficial mineralization processes could be extended in depth and enriched next to altered host units because of a nonlinear but self-organized distribution of geochemical- geophysical anomalies in Tekieh ore deposit region.

The Zarivar Lake, located in Kurdistan Province, is considered an important fresh water reservoir in the west of Iran. Recently, there has been an increasing risk of sediment and water pollution especially with heavy metals due to extended human activities in and around the lake. In this research, source of the metals Cu, Cr, V, Ni, Ni, Ti, Co, Sc, Zr, Li, Ca, Sr and Mg, present in the Zarivar Lake bed sediments, was determined using univariate (correlation matrix) and multivariate (Principle Component Analysis (PCA) and Cluster Analysis (CA)) statistical analyses. Thirty one samples were collected using a Grab Sampler. Based on CA results, the lake was divided into two northern and southern parts. Metals in the northern part were originated from a geological source. In the southern part, the mentioned metals had also a geogenic source except for Pb. Cluster analysis and PCA revealed an anthropogenic source for Pb in the southern part of the Zarivar Lake.

Granitic rocks of Malayer plutonic complex contain varieties of enclaves with different shapes, sizes, mineralogy and chemical composition. The interpretation of bivariant geochemical diagrams of major oxides and trace elements with respect to higher values of some of oxides such as MnO, TiO2, MgO, CaO & FeOt than host rocks in one groups of enclaves and moreover linear trend of these oxides and some of trace elements such as Ni, Cr, V indicate to different nature and mafic source of these enclaves (Mafic type) than host rocks and other enclaves (Felsic type).The study of chemical composition of this enclaves by using of univariant and bivariant statistical methods (bivariant regression analysis, correlation coefficients, cluster analysis and principle component analysis) indicate clear chemical contrast between mafic enclaves with felsic enclaves and granitic host rocks and in other side chemical affinity of felsic enclaves and their host rocks. Distinctive distribution of the majority of oxides and trace elements of mafic enclaves and host rocks and low values for R2 in regression analysis, low value of correlation coefficient of major element oxides and trace elements between enclaves and their host rocks, separate position of samples in cluster pattern and special direction of variants and samples of vectors in bivariant diagram of principle component analysis (PCA) are outputs of different geochemical characteristics of enclaves and host rocks. Moreover this correlates with different trends of each major oxides and trace elements in bivariant geochemical diagram (Harker diagram).

Landslides cause damage to property and pose a threat even to human lives. Suitable landslide susceptibility, hazard and risk models could help mitigate or even avoid the unwanted consequences resulted from such hill slope mass movements. For the purpose of landslide susceptibility assessment, a spatial database, which included 18 landslides prepared and analyzed using LISS-IV satellite imagery and geographical information system (GIS) technology. Analytical hierarchy process (AHP), the statistical index (Wi) and weighting factor (Wf) methods were applied to determine the significance of event-controlling parameters in triggering the landslides. The parameters included lithology, slope gradient, slope aspect, elevation, distance from faults, distance from stream and drainage density. These factors are different in geomorphic and geologic setting. The result showed that lithology, slope gradient and distance to stream play most important roles in landslide susceptibility. To confirm the practicality of the three susceptibility maps, there were compared with landslides activity map containing 12 active landslides. The results showed that the Wi method gave a more realistic picture of the actual distribution of landslide susceptibility, than the Wf method in the reservoir of Alborz dam.