No abstract

No abstract

Geochemical summary statistics for 18 elements in natural mateiials horn 117 landscape units have been compiled based on field and laboratory studies since 1958. Each landscape unit is brie f ly identified as to kind and location, and the expected concentration for one or more elements is gi\en together with factors indicating the degree of observed \ariation in the stuch and the degree of laboratory or "analytical" variation. Also listed are the observed lange and the total number of element analyses made in each study. The data on which these summaries are based have three attributes in common: They represent "large-scale" 01 legional geochemical studies; they represent background or "oidinary" natural geochemical \ariation; and they were collected according to objective sampling designs. The summaries clearly demonstrate the wide diveisity to be expected in elemental properties of landscape units and suggest that published element abundances for bioacl categoiies like1 "soil" or "carbonate rock" may be misleading. Fl F2 STATISTICAL STUDIES IN FIELD GEOCHEMISTRY One such study merits special mention. The U.S. Geological Survey has recently completed a geochemical survey of the State of Missouri, many aspects of which are unique to environmental geochemistry. It not only is a survey of a broad, geologically diverse area, but it was undertaken partly in support of active, trace-element related epidemiologic studies sponsored by the University of Missouri. Moreover, we think it is the first study of its kind in which an attempt has been made to characterize rocks, waters, soils, and plants chemically by a unified team approach using (and in part testing) efficient and objective sampling designs. Connor and others (1972) described this work in a preliminary way. Details of the study are available in a series of limited-distribution progress reports (U.S. Geological Survey, 1972a-f, 1973). The data tabulated in the present report represent the work of many people. Principal investigators are listed with a short description of each study, and authorship is cited for all published data. All unpublished data are preliminary. The reader is cautioned that some of the data summaries given here may be subject to minor revision. The sampling designs, data analyses, and geochemical summaries on which this report is based are statistical in nature; extended discussions of these subjects can be found in Miesch (1967a, b, 1972), Connor and others (1972), and Connor and Myers (1973). A proper list of acknowledgments for this report would comprise more than 100 people including computer programmers, specialists in data handling, and assistants in the field, laboratory, and office. Unquestionably, the most important contributors are the chemists, spectrographers, and other laboratory personnel who catalogued, prepared, and measured the concentrations of up to 69 elements in more than 8,000 samples of rocks, soils, and plant material over a period of more than 10 years.

Geochemical abundance of an element i•s regarded as the proportion of a total rock body, or group of rock 1 bodies, that is made up of the element and is equivalent to the population arithmetic mean of sample analyses, generally in units of percentage or parts per million. Computational problems encountered in estimation of abundances can arise from having limited ranges of analytical sensitivity, from having only small groups of data, and from reporting of data in broad geometric classes. 'l'hese problems can be partly .resolved by use of a combination of techniques described by Sichel (1952), Cohen (1959, 1961), and Krige (1960). The combination of techniques allows efficient estimation of geochemical a•bundances from a wide variety of frequency distributi'On types if the analytical discrimination is sufficient to allow effective use of data transformations. Examples of data •from the literature were used to demonstrate the techniques; up to 76 percent of the da•ta values were nrbitrarlly placed in the "not detected" class, and estimated arithmetic means agreed with those estimated from the complete data set to two significant figures. The types of frequency distributions displayed by the data •examples are generally representative of those commonly encountered in geochemical problems.

No abstract

No abstract

Major-element analyses of 497 composite samples of batholithic rocks (quartz diorites, granodiorites, and quartz monzonites) from the northern Peninsular Ranges and Transverse Ranges Provinces, southern California, form the basis for a mixing model that accounts for most of the compositional variation in the rocks. The compositional structure in the batholithic rocks as a group was found to be similar to that in the Sierra Nevada batholith, and indicates that four end members are sufficient to account for 85-97 percent of the variability in each of the eight major oxides. According to the model, the batholithic rocks formed from the mixing of basaltic and quartzofeldspathic end-member magmas, and the removal of variable proportions of plagioclase and mafic minerals, principally hornblende. Gabbroic rocks, common only in the western part of the region, could have formed from nearly uncontaminated magmas of the basaltic end member. Variations in the mixtures of basaltic and quartzofeldspathic magmas are presumed to reflect variations in their source materials at depth. According to the model, the compositions of the source materials do not vary smoothly over the region, but display a discontinuity along a line approximately coincident with the present San Jacinto fault zone. The discontinuity is roughly coincident with previously noted petrologic and isotopic discontinuities in the northern Peninsular Ranges and is interpreted as the western limit of significant contribution of continental materials to the batholithic rocks. The model gives no evidence of a discontinuity in the vicinity of the San Andreas fault zone. 'Average of eight analyses from Larsen and Draisin (1950, p. 69). 2Analysis 7 from Larsen and Draisin (1950, p. 71). ' See figure 3 for sampling localities and Appendix 1 for the analyses.

No abstract