Statistical Analysis Of Geochemical Data Research Articles

Exploratory multivariate statistical analysis of geochemical data from the Cenomanian–Turonian transition of the Yezo Supergroup, Hokkaido, Japan

The multivariate statistical analysis of major and minor elements and rare earth elements from 37 stratigraphical levels sampled across the Cenomanian–Turonian boundary in Hokkaido, Japan, yields evidence of substantial group differences in chemical properties of the sediments demarcated by the anoxic event at the transition between stages, both for the set of major and minor elements and for the set of rare earth elements. The multivariate statistical methods used were principal components, principal coordinates and canonical correlation, all of which were applied in a mode appropriate for compositional data. The results are an example of quantitative chemostratigraphy. A brief statistical overview of the main methods used is given.

Multivariate statistical analysis of geochemical data exemplified by Proterozoic dyke swarms in Sweden

The multivariate statistical analysis of chemical data in petrology and geochemistry entails principles that are little known among geological practitioners. Such data are compositional, and hence, lie in a sub-space of full space, the simplex. The concept of product-moment correlation of full space cannot be validly applied to them, because of the subcompositional incoherency of that statistic in the simplex. A second type of problem concerns the stability of multivariate estimates. Appropriate techniques, such as principal component cross-validation, are discussed exemplified and contrasted with inappropriate procedures using geochemical data on Swedish Proterozoic dyke swarms. Essential features of compositional covariance structure are exemplified in an Appendix.

Lithification of biogenic siliceous sediments: evidence from Neogene diatomaceous sequences of northeast Japan

Biogenic siliceous deposits are a significant component of Neogene stratigraphic sequences of the northeast Japan arc. Field observations demonstrate that the siliceous rocks are classified into five lithologic types: diatomite, opaline porcelanite, opaline chert, quartzose porcelanite, and black shale. Stratigraphic data suggest that the siliceous formations were deposited continuously from the Middle Miocene to Early Pliocene. Diatomite occurs throughout the sedimentary sequences and changes laterally into chert and porcelanite. Multivariate statistical analysis of geochemical data indicates that lithification and mineralogical transformation of silica-rich deposits was governed not only by compaction and cementation during burial but also by sorting of grains by biological sediment-seawater mixing. Diatomite is composed of frustules without fine ornamentations and delicate structures. Fine and delicate tests were more susceptible to dissolution than coarse and solid ones, and dissolved in the course of biological sediment disturbance. Ubiquitous occuerences of diatomite in the stratigraphic sequences are attributed to residual frustules which were more resistant to the diagenetic dissolution-precipitation reaction of silica. Biogenic silica is the main source of silica to form chert. Diagenetic precipitation of silica in diatomaceous beds through progressive sediment burial formed porcelanitic rocks, with opaline porcelanite being the precursor of quartzose porcelanite. Chert was not subjected to biological sediment reworking, and fragile shells were preserved during accumulation. Silicification of precursor deposits of chert was promoted by intraformational dissolution-precipitation reaction of silica with increasing sediment compaction and dehydration.

Assessing Differences in Composition between Low Metamorphic Grade Mudstones and High-Grade Schists Using Logratio Techniques

Based on stratigraphic and petrologic similarities, it has been proposed that rocks exposed in roof pendants of the central Sierra Nevada, California are part of a tectonically displaced fragment of the Cordilleran miogeocline. We have identified significant geochemical similarities between metamudstones making up this miogeoclinal fragment and rocks comprising its proposed parents in the Mojave Desert-southern Great Basin region by geochemical analysis of samples collected from Snow Lake and Boyden Cave roof pendants, and from the Precambrian to Cambrian section of the Cordilleran miogeocline exposed in the Nopah Range, southeastern California. In order to circumvent the constant-sum problem inherent in geochemical data, the data were transformed into a continuous-variable format using logratio techniques. When the differences in means of aluminum-normalized logratioed data derived from the Snow Lake-Boyden Cave data set and the Nopah Range samples were calculated on an element by element basis, ~86% of the differences were not statistically significant at the 95% confidence level. The large number of compositional similarities identified between Snow Lake, Boyden Cave, and Nopah Range samples are consistent with the idea that the metasedimentary rocks of the central Sierra Nevada batholith are composed of material shed from the western North American interior and that roof pendants in the central Sierra Nevada are part of a displaced miogeoclinal fragment. This work demonstrates that rigorous statistical analyses of geochemical data transformed into a continuous-variable format can be a useful tool in evaluating the plausibility of lithostratigraphic correlations between metamorphosed, complexly deformed, and displaced metasedimentary rocks.

Methodological contribution to geochemical subdivision of coal seams

The statistical analysis of geochemical data collected at Zasavje coal basin offered the possibility to use distribution characteristics of geochemical factors and clusters for subdivision and genetic interpretation of coal seams. In accordance to sampling scheme for variance analysis the interval samples of coal and bearing rocks brought data on ultimate, proximate and atomic emission spectrometric analysis, and silicate analysis of coal ash. Variability of all measured variables was studied on regional, vertical, sampling and analytical level. After factor analysis of 23 selected variables terrigenous (K, Cu, V, Cr, ash, Zn, Be, Co, Al2O3, SiO2), carbonate (Mn, CO2, Fe2O3, Ni), sulphide-sulphate (sulphide and sulphate sulphur, Fe2O3), nitrogen-hydrogen (O, N, H) and typomorphic factor (Sr, Ca, organic sulphur, C, Ba) have been determined. Five synthetic variables (factors) explain 79% of the total variance. Yet on the basis of the first three factor values the samples can be arranged by use of cluster analysis into sample groups (clusters), which represent various geochemical segments of coal seams. By using above presented methodological guidelines which are based on crosssections of individual coal-bearing basins structural units, the future multidisciplinary research can lead to genetic model of Slovenian coal basins. Presented article is a contribution to look after best possible geochemical methodology which purpose is to reach this aim.

Evolución geoquímica de las salmueras de las Lagunas de Monegros (Zaragoza, España) y análisis del equilibrio de la halita

Statistical analysis of geochemical data for brines from La Playa lake (Los Monegros, Zaragoza) has shown the existence of sampling-gaps during the whole evaporation process. But trends analysis with polinomial smoothing allow to recognize brine concentration paths and to verify the chemical divide concepto Physicochemical modelling with WATEQ4F and SOLMINEQ.88 for a Na + and a Cl - and halite saturation index determination shows the discrepancies among different approaches for activity coefficients calculations. The final conclusion is that only Pitzer formulation is adequate for this high ionic force waters, but in this case, D extended Debye-Huckel equation is not excessively defficient.

The hydrothermal component in ferromanganese nodules from the southeast Pacific Ocean

Statistical analyses of geochemical data from ferromanganese nodule samples representing a broad area of the southeast Pacific Ocean indicate nodule compositions are controlled by four factors: (1) suboxic diagenesis; (2) oxic diagenesis; (3) hydrogenous precipitation; and (4) hydrothermal precipitation. The first three are identical to those identified in studies of nodules from the North Pacific; the fourth factor represents the first quantitative determination of a significant hydrothermal end-member in ferromanganese nodules. Nodules apparently acquire a hydrothermal component through the accretion of colloidal-sized hydrothermal precipitates which are widely dispersed from their origin at vents along the East Pacific Rise (EPR). Nodules recovered from beneath the “shadow” of the present-day hydrothermal plumes emanating from the EPR contain significantly higher amounts of hydrothermal material relative to nodules from other areas.

日本の第四紀火山岩の地球化学データの管理と統計解析

日本の第四紀火山岩の地球化学データの管理と統計解析