Rollinson Hugh R. Using geochemical data- evalution, presentation, soundofheaven.info - Ebook download as PDF File .pdf), Text File .txt) or read book online. H. R. Rollinson. Using Geochemical Data: Evaluation, Presentation, Interpretation. London (Longman Scientific and Technical), xxvi +. Using Geochemical Data: Evaluation, Presentataion, Interpretation. Naiyar Imam clicking the 'Download' button above. READ PAPER. Download pdf. ×Close.
|Language:||English, Spanish, Indonesian|
|ePub File Size:||22.51 MB|
|PDF File Size:||14.21 MB|
|Distribution:||Free* [*Regsitration Required]|
PDF | On Jan 1, , Hugh R. Rollinson and others published Using Geochemical Data: Evolution, Presentation, Interpretation. PDF | On Jan 1, , H.R. Rollinson and others published Using geochemical data: Evaluation, presentation, interpretation. Using Geochemical Data brings together in one volume a wide range of ideas ByHugh R. Rollinson DownloadPDF MB Read online.
Furthermore, the isotopic evidence indicates that the different reservoirs have made their contributions to the lava pile at different times in the history of the volcano. Bell K. Dostal J. J ames, R. Mass Spectrometry, Acta, 46,
For example, in the list of analyses of tonalitic and trondhjemitic gneisses in Table 2. This is illustrated in Figure 2. This is traditionally done using regression analysis. Regression analysis is the subject of number of statistical texts e. Below some of the more popular forms of regression aredescribed. Unfortunately, it is often not appropriate. Some of these alternatives are reviewed below.
Thus, unlike ordinary least squares and least normal squares regression, the slope of the reduced major axis line is independent of the correlation coefficient r. In Figure 2. The equation for each line is given. This is necessary when some data points are less reliable than otners and so are more subject to error. The regression lines are: In robust regression, outlying values are downweighted.
The statistical difficulties resulting from this feature of geochemical data are formidable and are discussed below in Section 2. The exception is in geochronology, and this is discussed in Section 2.
Pearce diagrams were originally developed to avoid the effects of closure inherent in plotting percentages, the conventional method of displaying major element geochemical data. Thus different slopes identify different mass transfer processes.
However, this argument too is flawed, for regression lines drawn through the data have incorrect slopes. This is because, in the case of ordinary least squares regression, the slope of the line is directly related to the correlation coefficient.
Eqn [2. Small variations in the concentration of the ratioing element give rise to large and similar variations in both ratios of the Pearce plot. Butler pointed ou. In brief, the problem is as follows. Thus, components of percentage data are not free to vary independently. In terms of correlation theory the usual criterion of unrelatedness or independence does not hold, nor are the variances and covariances of the data-set independent.
In addition, subcompositions have variances which show different rank ' orderings from those in the parent data-set. Consider the data presented in Table 2. This observation has important implications for approaches such as discriminant analysis, frequently used in geochemistry see Section 2. The problem of closure is not removed when the data is transformed into cations Butler, nor is it removed when the problem is presented in graphical rather than statistical form.
The two are the same inasmuch as both require the taking of ratios, but there the similarity ceases. The covariance structure of log-ratios is free from the problems of the negative bias and of subcompositions which bedevil percentage data.
The choice of variable as the divisor is immaterial because it is the structure of the matrix which is of importance rather than the individual values of the covariances. The final form of compositional covariance structure is centred log-ratio covariance matrix.
These were chosen for their apparent geological simplicity. Large negative values in the two covariance matrices tend to confirm the variability indicated in the variation matrix. Plutonic rocks are more problematic. The device which is most commonly used and has proved invaluable in the examination of geochemical data is the variation diagram.
Below we consider some of the more important mixing processes. The data are given in Table 3. The second application is to mudstones. The major and trace element chemistry of modern muds reflects the degree of weathering in their source Nesbitt Variation diagrams 71 Advanced weathering trend Weathering trend ; The advanced weathering trend for granite is also shown. Compositions are plotted as molar proportions and the compositions of plagioclase, K-feldspar, muscovite and kaolinite are shown.
There are two possible explanations of such trends. Table 3. Sediments are not so well studied although chemical changes during the diagenesis of sandstone-clay sequence are well described and reflect the breakdown, with progressive burial of orthoclase and plagioclase and the conversion of smectite to illite.
For example, if in the case. These are: Triangular variation diagrams are used when it is necessary to show simultaneous between three variables. Variation diagrams 75 The plotting procedure for triangular diagrams is illustrated in Figure 3. The values are plotted as follows: The point at which the two lines intersect is the plotting position.
The AFM diagram is most commonly used to distinguish between tholeiitic and calc-alkaline differentiation trends in the subalkaline magma series.
The coordinates for the boundary lines are given in the caption to Figure 3. Examples of the trends characteristic of the tholeiitic and calc-alkaline rock series are also plotted in Figure 3. This is more difficult to quantify. The method is illustrated in Figure 3. Back-projection shows that five of the elements converge and reduce to zero at Figure 3. Inflections are most obvious where the number of fractionating minerals is small, such as in basaltic melts. In calc-alkaline volcanic rocks, where the number of.
In highly porphyritic volcanic rocks t: The differences between the calculated composition and the actual composition of the Lolo basalt, using the mixing program of Wright and Doherty , suggest that the solution is acceptable.
Hanson and Langmuir and Langmuir and Hanson modelled basaltic systems from single-element and single-component mineral-melt distribution coefficients. These are combined with mass balance considerations and the stoichiometry of the mineral phases to calculate phase equilibria. Particularly interesting is their model for the partial melting of mantle pyrolite at 1 atmosphere pressure. Using the equations of Roeder and Emslie for the partitioning of magnesium and iron between olivine and melt, they calculated the abundances of MgO and FeO in the resultant melts and residual solids.
The diagram in Figure 3. The parent composition is where the two fields meet. Phase diagrams of this type serve two useful functions. The aim of this section therefOl'e is to describe some of the main diagl'ams and projection schemes used in plotting experimental and natul'al rock data. Tuttle and Bowen and subsequent workers have determined the compositions at which the phases quartz, orthoclase and albite. The plotting procedure requires three steps: At approximately 3.
At 30 kb the assemblage is coesite - sanidine hydrate - jadeiite. The reasons for this are outlined below. Firstly, in most experimental investigations compositions are projected from HzO onto the plane Ab-Or-Qand it is assumed that the melt is water-saturated. Thirdly, it is important to know whether the bulk compositions sampled represent igneous liquid compositions or whether they are in part crystal cumulates.
We look at the effects of reducing the water content of the melt and adding anorthite to the melt, thus extending the applicability of this system to granodiorites and tonalites.
The best data are those of Steiner et al. Luth has estimated the position of the 10 kb dry minimum and Huang and Wyllie have estimated the position of the 30 kb dry quartz-alkali feldspar field boundary. The addition of anorthite to the 'granite' system shifts compositions into the granodiorite and tonalite fields.
These data are presented in Figure 3.
The sources of data and the plotting positions are listed in Table 3. However, at best, these experimental data lend themselves to the qualititative interpretation of natural alkaline rocks.
The most used projections are: The olivine projection is illustrated in Figure 3. The olivine-gabbro plane divides the diagram into nepheline normative compositions enstatite-poor and tholeiitic compositions enstatite-rich.
The inset shows the relative positions of olivine and the plane of projection in the olivine projection of the CMAS system. Various pro;ection procedures are in use. Presnall et al. Walker et al. In this projection, therefore, plagioclase compositions are spread along the silica-anorthite edge of the Di-Ol-An-: It was presented in Section 3.
The phase diagram is useful for estimating thephases present in the initial stages of low-pressure crystallization of given tholeiite. This is also true for the algorithm of Elthon Thus Presnall et al. Grove et al. The projection scheme is similar to that of Walker et al. The procedure is as follows: The 1 atm liquidus boundary curve is also shown.
Rather they point the way to less probable and more probable options. Using trace element data 4. Particularly important is the fact that there are mathematical models to describe trace element distributions which allow the quantitative testing of petrological hypotheses.
In this chapter we first develop some of the theory behind the distribution of trace elements and explain the physicallaws used in trace element modelling. Several groups of elements in the periodic table are of particular geochemical interest Figure 4. The most obvious in this respect are the elements with atomic numbers 57 to 71, the lanthanides or rare earth elernents REE as they are usually called in geochemistry. In geochemistry, this latter term is usually restricted to the first transition series and includes two major elements, Fe and Mn.
The elements in each of these respective groups have similar chemical properties and for this reason are expected to show similar geochemical behaviour. In detail there are degrees of compatibilty and incompatibility and trace elements will vary in their behaviour in melts of different composition.
Figure 4. Elements with the same ionic charge and size are expected to show very similar geochemical behaviour. Their very low concentrations, however, lead to relatively simple relationships between composition and activity.
This states that at equilibrium the activity of trace element is directly proportional to its composition: In the case where trace elements form the essential structural constituent of minor phase, such as Zr in zircon, Hehry's Law behaviour does not strictly apply. Compatible elements are placed towards the bottom, left-hand corner of the diagram.
Some of the first transition series metals transition elements and the PGE elements, are quoted for six-fold coordination. Philpotts and Schnetzler, Dunn, However, we are not always in the fortunate position of having as much information available as this and it is often necessary to 'make do' with the available data. It is for this reason that the partition coefficients listed in Tables 4. The composition control of minerallmelt partition coefficients between the REE and hornblende is illustrated in Figure 4.
Two experimental studies, published at the same time, seem to show conflicting results. Data from Tables 4. However', Green and Pearson showed that in the case of the partitioning of the REE between sphene and silicate liquids the water content of the melt 0.
In Table 4. The compilation in Table 4. The several sets of REE data shown are generally in good agreement with each other.
Colson et al. Table 4. REE partition coefficients for hornblende are higher in andesites than basalts, but are comparable between andesites and basaltic andesites. Partition coefficients for the REE in any one of the ferromagnesian minerals are variable. In the case of pyroxenes the light REE values of Mahood and Hildreth are probably in error and the alternative values of Michael are -used here. Data from Table 4. At present our knowledge is patchy, for some processes are well understood and there are mathematical models available to describe them.
Other areas, equally important but less amenable to the quantitative approach, are understood qualitatively. Equilibrium is therefore only achieved between the melt and the surfaces of mineral grains in the source region.
It is worth noting in passing that physical models of melt extraction describe melt fractions in ter. This formulation of the batch melting equation is very straightforward to use.
In the case of modal melting i. Taking the simple case where D is calculated for the unmelted residue Eqn [4. When D is small, expression [4. Enrichment -and depletion in the solid residue in equilibrium with the melt Eqn [4.
Compatible elements, however, at small degrees of melting remain very close to their initial concentrations. The shaded region is the area in which enrichment is impossible.
In the case of modal melting Eqn [4. Rayleigh fractionation describes the extreme case where crystals are effectively removed from the melt the instant they have formed. The equation for Rayleigh fractionation is [4. In this case small melt fractions are removed instantly from the source but aggregate together. This process produces very similar results to that of batch melting. Rayleigh fractionation is less effective than batch melting in changing the ratio of two incompatible elements, for the curves for 0.
The solidification zone progressively moves through the magma chamber until crystallization is complete.
Equation [4. At low values offthe enrichment of incompatible elements and the depletion of compatible elements are not as extreme during in situ fractional crystallization as in Rayleigh fractionation. Figures 4. Compatible elements are strongly depeleted. In reality it is likely that melts are neither instantaneously removed from the source nor do they remain totally immobile in their source. The numerical effects of zone refining are illustrated in Figure 4.
Langmuir et al. At higher values of the degree of enrichment of incompatible elements is reduced. Other elements are more soluble. This is illustrated for ions in eight-fold coordination state in Table 4. It is this phenomenon which is used in geochemistry to probe into the genesis of rock suites and unravel petrological processes.
Composite Composite Leedey Composite Avg. CI Avg. Leedey chondrite. However, the concentrations of the REE in the solar system are very variable because of the different stabilities of the atomic nuclei. This pattern of abundances is also found in natural samples. Chondritic normalization therefore has two important functions. This is sometimes referred to as the Masuda-Coryell diagram after the original proponents of the diagram Masuda, ; Coryell et al.
The variability I? The patterns show both variety in shape and in concentration range. Elements with even atomic numbers have higher abundances than those with odd atomic numbers. In fact the two sets of values are very similar and lie in the middle of the range of values currently in use. This 'average sediment' is often used as the normalizing value for REE concentrations in sedimentary rocks.
This assumes that sedimentary processes homogenize the REE previously fractionated during the formation of igneous rocks. Interpreting REE patterns The REE are regarded as amongst the least soluble trace elements and are relatively immobile during low-grade metamorphism, weathering and hydrothermal altera- tion.
River Luce, Scotland; 0. Lm filter. In Section 4. This is evident from the partition coefficients plotted in Figure 4. The REE are compatible in hornblende in felsic and intermediate liquids and the highest partition coefficients are between Dy and Er.
In basaltic liquids the partition coefficient for Lu is more than times greater than that for La. The effect is less extreme, although still large, in felsic liquids. The REE c6ntents of rivers and seawater are extremely low Table 4.
In comparison, the effects of weathering and diagenesis are minor. The data are given in Table 4. Normalizing values are from Table 4.
They are an extension of the more familiar chondrite-normalized REE diagrams in which other trace elements are added to the traditional REE diagram. The terms 'mantle or chondrite -normalized multi-element diagram' or 'incompatible element diagram' do not roll off the tongue with ease and the more colloquial 'spider diagrarn' or 'spidergram' for an individual pattern is used here. These include an estimated primitive mantle composition and chondritic meteorites two 'views' of the primitive undifferentiated earth.
This state affairs is not satisfactory and some standardization is desirable. First, however, we consider the present 'state of the art'. The primitive mantle is the composition of the mantle before the continental crust formed. One of the most frequently used estimates of its composition is that of Wood et al. The order of elements Figure 4.
Thompson et al. The immobile elements are arranged from right to left in order of increasing incompatibility Figures 4. Normalized multi-element diagrams For igneous rocks two multi- element diagrams are sufficient: The most commonly used normalizing values are those for average shale such as average post-Archaean shale and the North American shale composite NASC , representing 'average crustal material', although average upper continental crust is also used.
Trace elements in marbles and calc-silicate rocks were normalized to average Phanerozoic limestone Condie et al. Gold is often associated with the latter group. In rocks such as ocean-floor basalts PGE concentrations are so low that some elements are below the limit of detection. Normalizing values currently in use are given in Table 4. The analogy with the REE is not close, however, because the PGEs are not ordered from light to heavy in keeping with their order in the periodic table.
The elements are broadly arranged in order of increasing compatibility in the primitive mantle from left to right across the diagram. Primitive mantle values are listed in Table 4. There is some debate over the significance of the mantle concentrations of PGEs. They range between 0. Archaean mantle. The normalizing values are taken from Langmuir et al. Ti anomalies indicate the role of Fe-Ti oxides. The normalizing values are taken from Sun - Table 4.
Variation diagrams are discussed in detail in Section 3. The task for the Incompatible element plots Bivariate trace element plots Most fruitful are trace elements which show extreme behaviour, such as the highly and the highly compatible elements. Incompatible element concentrations are particularly senslt1ve to partial melting processes see for example Figures 4.
This is true for both batch melting and fractional melting but is most extreme in fractional melting. Incompatible element concentrations also vary during fractional crystallization, although the effect is most strongly marked in AFC processes.
In the case of mantle melting, the following groups of elements have almost identical bulk partition coefficients during mantle melting: Secondly, this method is difficult to apply to granitic rocks. Ratio-ratio plots do have some inherent problems and the reader should take note of possible spurious correlations arising from the common denominator effect discussed in Section 2.
Compatible Compatible trace element concentrations change dramatically in an igneous liquid element plots during fractional crystallization Figure 4. Thus bivariate plots of compatible elements, plotted against an index of fractionation e. Hildreth compares the relative concentrations of the early and late members of the Bishop's tuff Figure 4. Data from Hildreth Below we describe these two modes of presentation in some detail. These are illustrated in Figure 4. Partial melting vectors are used to show changing melt and source compositions during the partial melting of given source composition and mineralogy.
The effects of different melting models, source 70 70 10 Figure 4. The direction of the lines shows the compositional change in the residual liquid when the specified phase is progressively removed during fractional crystallization. Details of the calculation are given in Table 4.
This means that it is possible to view trends in the data.
The process is illustrated in Figures 4. The vector for modal batch meltin between 0. The partition coefficients were taken from Table 4. For example, Thompson et al. Details of the calculations are given in Tables 4. Multivariate diagrams are used to compare calculated and measured rock compositions..
Arth estimated the mineralogy of the fractionating assem- blage from the proportions of phenocrysts present in the lavas and calculated REE patterns which show excellent agreement with the observed REE patterns in the rocks. Similarly, Condie and Crow Modelling trace element processes in igneous rocks. Curves are shown for 0. The calculations were made for the ten REE shown using Eqn [4. The equations of Langmuir et al. The first step in using the inverse approach to the study of trace elements is to identify the likely physical process which accounts for the variation in the data.
For example, elements which are compatible will vary drastically in concentration during fractional crystallizsation, whilst highly incompatible elements will vary most in abundance during partial melting Minster and Allegre, In this case the unknowns are 1 the initial concentration of the trace elements in the parent magma, 2 the bulk partition coefficients for the elements and 3 the degree of crystallization corresponding to each sample. The initial concentrations of trace elements in the melt were estimated using Ni.
Bulk partition coefficients were calculated using the method of Allegre et al. Minster et al. In the case of partial melting the unknowns are l the chemistry of the source, 2 the bulk partition coefficient for each element considered and 3 the degree of partial melting for each sample.
Bender et al. The results of preliminary calculation are inspected and refined as necessary. Chapter 6 Using radiogen,ic isotope data 6. Historically they were first used to determine the age of rocks and minerals. The former application is normally described as geochronology, the latter as isotope geology or isotope geochemistry. In the first part of this chapter the main principles of geochronology are briefly described and the interpretation of geochronological results are reviewed.
They showed that the process of radioactive decay is exponential and independent of chemical or physical conditions. Defines age of earth as 4. The precise measurement of absolute isotope concentrations is difficult; instead, isotope ratios are normally determined. In the case of the Rb-Sr isotope system the ratioing isotope is 86Sr and Eqn [6. Since Eqn [6. The methodology is illustrated in Figure 6. South Africa from Hamilton et al. Precise results are obtained from statistical line-fitting procedures which estimate the slope and intercept of the isochron.
If the MSWD is greater than 2. This is illustrated in Figure 6. For this reason model ages for the continental crust are usually calculated with reference to the depleted mantle DM reservoir rather than CHUR.
It is important when calculating model ages to remember the assumptions upon which they are based, for these are not always fulfilled. Firstly, assumptions are made about the isotopic composition of the reservoir which is being sampled - either CHUR or depleted mantle.
This aspect of model age calculations in itself raises three further problems. The difficulty has arisen because the Sm-Nd technique evolved very rapidly and different laboratories developed different normalization schemes in parallel.
These are summarized in Table 6. In this case there are two possible solutions, depending upon which mantle model is preferred. Table 6. Different minerals will close at different temperatures and different isotopic systems in the same mineral will also close at different temperatures. An Sm-Nd whole-rock isochron for this suite gives an age of 2. Calculations made using Eqn [6.
The departure from the 1: Thus Rb-Sr isochrons are rarely useful in constraining crust formation ages. Blocking temperatures for isotopic systems in minerals are illustrated in Figure 6. Reuter and Dallmeyer dated the timing of cleavage formation in 10w-grade pelites using this method. Recently, the application of the ion microprobe to the U-Pb isotopic analysis of single zircon crystals Compston et al.
The U-Pb method is also applicable to the minerals sphene, monazite and epidote. In the case of mantle-derived granites, the model age gives the time of mantle fractionation source 1. The model ages have little real meaning for they neither reflect the crystallization age of the granite nor the age of the crustal source Arndt and Goldstein, This simple observation has led to two important developments in isotope geochemistry.
The time-scale of these processes is short and equivalent to the time-scale of most familiar geological processes. The second.
For example, we will show below that there are several sources for oceanic basalts in the upper mantle. It is of some importance to discover how and when these separate sources acquired their separate identities. For example, Rb is the element most concentrated in the crust relative to the depleted mantle whereas Sr, Sm and Lu are the least concentrated. Sm-Nd are immobile under hydrothermal conditions and so their isotopic composition reflects the actual proportions of rock or magma involved in specific petrological processes.
The difference in behaviour between the different isotopes of lead allows the identification of several isotopic reservoirs Table 6. Strontium is relatively immobile under hydrothermal conditions, although Rb is more mobile.
Sr therefore reflects fairly closely the original bulk composition of suite of rocks, and Rb less so. In addition the Rb-Sr system shows the most extreme differences in incompatability between the parent and daughter elements. Rb and Sr are easily separated, so that there is extreme fractionation between crust and mantle leading to the accelerated strontium isotope evolution of the continental crust relative to the mantle see Figure 6.
The composition of each of these sources is summarized in Table 6.
In addition Table 6. In the following section each of the important mantle and crustal reservoirs is described and its particular isotopic character highlighted. Central 0. Australia 0. Africa 0. Britain 0. Phanerozoic Beni Bousera 0. Oceanic mantle Young magmatic rocks record the isotopic composIt1on of their source directly.
Their possible location in the mantle is given in the cartoon in Figure 6. The end-member compositions are as follows. The enrichment is thought to have taken place between 1. Most non-enriched mantle reservoirs plot in the upper left 'depleted' quadrant cf. Figure 6. The equations are J 00 HIMU 0. Furthermore, the isotopic evidence indicates that the different reservoirs have made their contributions to the lava pile at different times in the history of the volcano.
Th levels are lower than those in the upper crust but not as depleted as uranium. The lithosphere beneath Proterozoic mobile belts, however, more closely resembles the depleted mantle found beneath older ocean basins Menzies, However, the strontium isotopic composition of seawater has changed with time.
For instance, over the total history of the Earth Figure 6. Flegal et al. Diagrams of this type are particularly useful for depicting the isotopic evolution of crustal rocks of differing ages and with different parentl daughter isotope ratios. The starting point for the evolution of strontium isotopes is the 87 Sr I 86Sr ratio at the formation of the Earth.
Measured values in present-day basalts from Samoa White and Hofmann, scatter about the 10wer of these two values. Samoa White and Hofmann Figure 6. For example, it is easy to see from Figure 6.
Radiogenic isotopes in petrogenesis 0. J 00 Initial ratio 0. The compositions of CHUR for the present day and at 4. Currently there are several models for this depleted mantle source and these are portrayed relative to CHUR the epsilon notation in Figure 6. The main differences between the different models of the depleted mantle are 1 the time at which the depleted mantle differentiated from the bulk Earth and 2 whether the depletion of the mantle was linear or whether it varied with time.
Liew and McCulloch proposed that the light REE depletion from the chondritic reservoir took place between 2. Goldstein et al. Friend Reviews. To see what your friends thought of this book, please sign up.
To ask other readers questions about Using Geochemical Data , please sign up. Be the first to ask a question about Using Geochemical Data. Lists with This Book. This book is not yet featured on Listopia. Community Reviews. Showing Rating details. Sort order. Oct 23, Jairo Conde rated it it was amazing. It is important for my investigation.
Natalie Starkey rated it it was amazing Feb 01, Jairo Conde rated it it was amazing Aug 22, John Friis rated it liked it Nov 16, Laurence rated it really liked it Mar 14, Richard rated it it was amazing Apr 21, Amitav rated it liked it Dec 11, Sonia Andrea rated it it was amazing Mar 13, Mohamed Idbaroud rated it did not like it Dec 23, Didem rated it it was amazing Nov 27, Nika Maisuradze rated it it was amazing Nov 22, Simobellam rated it it was amazing Dec 20, Marcos Jauregui rated it it was amazing Jul 22, Kristian rated it really liked it Apr 22, Saeed rated it liked it Jun 06, CMS rated it liked it Feb 21, Adalberto rated it liked it Nov 05, Andy rated it liked it Jan 14, Quinox rated it liked it May 20, Selection Methods 3.
Controls and Sources of Errors 4. Analysis of Geochemical Data 5. Using Major Element Data 6. Using Trace Elements Data 7. Using Radiogenic Isotopes and Isotope Data. Nielsen Book Data Publisher's Summary Using Geochemical Data brings together in one volume a wide range of ideas and methods currently used in geochemistry, providing a foundation of knowledge from which the reader can interpret, evaluate and present geochemical data.
Nielsen Book Data Subject Geochemistry. Bibliographic information. Browse related items Start at call number: R75 Librarian view Catkey: