Blake, S., Ivey, G.N., 1986. Magma-mixing and the ... 37. Cox, K.G., 1978. Komatiites and other high-magnesia lavas: some problems. Phil. Trans. Roy. Soc.
Factors of Pre-Eruptive Magma Evolution in the Continental Destructive Zones by Boris I.Malyuk
1989, Institute of Geology & Geochemistry of Combustible Minerals, Lviv, Ukraine Preprint No. 89-1. – 56 p. (In Russian)
Contents INTRODUCTION .......................................................................................................1 1. DEFINITION AND BASIC FEATURES OF TECTONICS AND MAGMATISM OF THE CONTINENTAL DESTRUCTIVE ZONES ..........................................................................2 2. STAGES OF THE PRE-ERUPTIVE MAGMA EVOLUTION .................................................2 2.1. FOCUS (melting zone) STAGE...........................................................................3 2.2. TRANSIT STAGE .............................................................................................4 2.3. CHAMBER STAGE.......................................................................................... 11 3. PROBLEM OF THE RECONSTRUCTION OF THE PRE-ERUPTIVE MAGMA EVOLUTION IN THE CONTINENTAL DESTRUCTIVE ZONES ........................................................................ 21 CONCLUSIONS....................................................................................................... 24 REFERENCES ......................................................................................................... 25
INTRODUCTION In the recent decades the geology has got a significant progress in the studies of deep structure and evolution of the under-crustal zones of the Earth. One of the brightest results of these works is the independent scientific direction establishment, the Geodynamics, where the whole movements of substance inside the planet and the driving forces of the variety of these processes are being studied. Geodynamic investigations engage two basic approaches: • •
geophysical study of the deep horizons of the crust and mantle with the subsequent mathematical simulation of the tectonosphere blocks movement the petrological study of magmatic associations for the purpose of the reconstruction of the conditions for their formation
In the first case mainly structural features are being taken into account whereas in the second one the matter of deep zones is subjected to the investigations. Therefore both approaches are not alternative, supplementing each other during comprehensive studies. The modern arsenal of petrological studies includes a group of the precise methods of the magmatic rock analysis, in particular determination of the minor, trace and REE element contents, and also important isotopic ratios (isotopes of lead, strontium, neodymium). These methods create a reliable basis for the interpretation of the fine geochemical features of the magmatic formations, many of which can have indicator value from the point of view of geodynamics. In this connection much depends on the proper geological interpretation of the petrological and geochemical features. This systematic problem remains to be actual one, since examples of incorrect interpretation of the geological records are known from publications. In particular, the features of the magmatic rocks available for the direct studies (i.e. samples located in the upper parts of the Earth's crust) are being extended sometimes also upon the upper mantle without proper accounting of the modifications, which the magmas could have experienced in the process of the motion from their origin sites to the surface. The influence of these modifications, most important of which is considered to be the fractional crystallisation, is discussed in petrology beginning from the classical works of H.Rosenbusch, R.Daly, F.Levinson-Lessing, N.Bowen. The intensive development of 1
questions of the pre-eruptive modification of magmas was extended in the recent two decades on the basis of the wide use of experimental studies and analytical methods of the solution of the problems of thermodynamics and fluid mechanics. At the present time the specific prerequisites are created for the correction of the basic principles of the interconnection of the magmatism and geodynamics taking into account the factors of the pre-eruptive magma evolution. It is natural that in each geodynamic situation this question requires individual approach. In this work a generalised analysis of the factors of the preeruptive magma evolution is carried out with respect to the continental destructive zones. Besides the indicated geodynamic aspects these zones also provide the interest from a metallogenic point of view, as the places of extraction of a number of economically important elements from the deep zones.
1. DEFINITION AND BASIC FEATURES OF TECTONICS AND MAGMATISM OF THE CONTINENTAL DESTRUCTIVE ZONES Continental regions consist of large number of structures, which were developed in the different geodynamic situations. Among them, however, it is possible to distinguish the group of the elements, which were being formed exclusively on continental crust. This includes first of all the rift zones of epi-platform and epi-orogenic types [10, 11], and also diverse trough structures and deep-seated faults. Since one of the basic geodynamic elements of their evolution is the crustal breakdown up to its complete destruction and structure re-building, they can be united into one group under the conditional term "Continental Destructive Zones" (hereafter CDZ). Formation and evolution of the CDZ can be appropriately interpreted from the positions of the rifting model. At present are examined two basic rifting mechanisms: active and passive [12, 49, etc.]. The mechanism of active rifting assumes the destruction of the upper part of the lithosphere under the effect of an ascending mantle diapirs (asthenoliths) that have the upper or even lower mantle origin. In contrast to this, the passive rifting is developed as a result of regional tangential tensile stress impact on the different-scale blocks of the lithosphere. The uplift of the asthenosphere roof in passive rifting is a consequence, but not by the reason for the tension of lithosphere. These circumstances are quite significant for moving the magmatic melts and, therefore, they can in a specific manner affect the factors of the pre-eruptive magma evolution. The defined categories of the continental structures, included into the CDZ class, outline rather wide spectrum of magmatic associations. Strictly rift zones are characterised by the wide development of the alkaline magmatism, although for some rifts tholeiitic and calcalkaline series are characteristic [6, 7, 10, 11]. Alkaline affinity is inherent also in many deep-seated fault zones confined to the world rift system, where kimberlite, carbonatite, and lamproite rocks occur. Finally, trough structures, in particular lower-Precambrian greenstone belts, are characterised by the domination of the magnesium and ultramagnesium magmatism of normal alkalinity. Thus, CDZ contain distinct variety of magmatic records, and respectively the analysis of the factors of the pre-eruptive magma evolution must be achieved over a wide range of the parameters taking into account two basic rifting mechanisms.
2. STAGES OF THE PRE-ERUPTIVE MAGMA EVOLUTION Obviously the pre-eruptive magma evolution can be maintained within a solidus-liquidus interval only and, therefore, from the onset of the first drops of melt in the focus of magma formation up to the contact of melt with the atmosphere or the hydrosphere. Hence, it looks reasonable to distinguish several stages of the pre-eruptive magma evolution in CDZ, which reflect differences between both the physical and chemical characteristics of magmas, and the external physico-mechanical and tectonic conditions at the different stages of generation and transport of the melts. Based of the general petrological and tectonic considerations, it is possible to distinduish three main stages of the pre-eruptive magma 2
evolution, which will be examined for the CDZ case, but, apparently, they are applicable also to other geodynamic situations: 1. the focus (melting zone) stage, which corresponds to the stage of the magma generation 2. the transit stage, which corresponds to the stage of the magma ascent 3. the chamber stage, which characterises a set of the magma conservation events in the transitional reservoirs in between their uplift from the generation foci to the surface. The ideal version of the magma evolution does assume the simple sequence of stages from 1 through 2 to 3. However, under the actual conditions the history of the development of magmatic process can be considerably more complex. Apparently, the common complication of situation is multiple repetition of the sequence of stages from 2 to 3, indicating existence of the multi-level magmatic column, which includes a series of the transitional chambers, connected by the channels of the magma transport. The complex versions of the combination of stages 2 and 3 can cause the exceptional poly-directivity of the deep melt evolution, whose potential scales will become visible from the following presentation.
2.1. FOCUS (melting zone) STAGE The general tendency of the magma evolution at the focus stage is generation of some volume of the melt. Therefore the main factors of the magma evolution should be considered as temperature increase in the melting zone, and a pressure drop within its limits. The magma evolution includes both the directed modification of the melt composition and change in their structural position in the surrounding solid framework. The process of partial melting that controls melt composition is well described in the literature; therefore there is no need to pay any special attention to its basic moments. Practically always the composition of partial melts it going to be considerably less in magnesium content than the composition of source rocks (substratum). But sometimes in the melting of peridotites at pressures ~ 150 kbar under the dry conditions the partial melts (komatiite liquids) are very close in composition to substratum, i.e., melting has an eutectoid nature [68]. In all other cases local equilibrium during partial melting corresponds to some specific level of the melt magnesium content. Variations in the melt compositions, besides noted important P-T parameters, are determined also to a considerable degree by a quantity of fluid and by its species composition. For instance, it was experimentally shown that the substantially aqueous fluids do cause an increase in the contents of silica in the melts, whereas carbon dioxide fluids lead to the alkali content increasing [1]. Recently appear the data about the important petrogenetic role also of other volatile components (phosphorus, chlorine, fluorine, etc.). Probably, to the greatest degree the role of fluids is occurred in the formation of the kimberlite and close to them lamproite and carbonatite magmas, which are, in the opinion of I.D.Ryabchikov [14], the fluid-melt mixtures generated in the zones of extensive mantle metasomatism. It is obvious that in this case the melt composition bears the least defined information about the composition of substratum, since the magma formation to a considerable degree is controlled by the introduction of the fluid, which actively modifies the composition of both the melt and refractory residium. Ideas about the composition of upper mantle in the alkaline magmatic provinces are usually based on the results of studying the xenoliths, carried by these magmas, although some researchers [9] bring into question the indicator role of xenoliths as the upper mantle remnants. Perhaps, in most extent the conditions of partial melting at the focus stage and the composition of the magma-forming substrata reflect the magmas, which were being generated during the equilibrium partial melting with the participation of the limited quantities of fluid. In this, probably most common case the high value for the magma evolution has the spatial distribution of the gradients of the temperature and pressure in the focus of magma formation. The zone of partial melting can be connected with the local thermal anomaly in the upper mantle or with the decompression in the deep fault zone. M.O'Hara [58] has showed that this zone has lenticular form and it must be thinned out to the periphery, which implies the decrease of the degree of melting substratum and 3
respectively a change in the composition of partial melts to the periphery of the zone. Thus, in the zone of partial melting simultaneously can be generated the melts, which correspond to different equilibrium conditions with the substratum. The gross composition of melt in the focus will correspond to a certain average value of conditions for equilibrium melting and therefore it can be examined only as the integrated index of the process of magma formation. Similar influence will, apparently, render such the factor of magma formation as the compositional heterogeneity of the source regions. In particular, many recent models of the structure of the upper mantle [2, 3, 23-25] do assume significant role in the composition of mantle substratum besides peridotites also of diverse pyroxenites and eclogites varieties; their relationships are rather well described by cake-layered-and-veined model [1]. From this point of view it is possible to assume that in the focus of magma formation the species of substratum with different temperatures and melting conditions will be present. The respectively resulting gross composition of melt in the focus, just as in the foregoing example, will correspond to some average conditions of melting in the compositionallyheterogeneous substratum, and it also should be considered as the averaged index of the process of magma formation. The focus stage of the evolution of the magmas just in CDZ is characterised by the fact that the geodynamic situation in the maximum degree contributes to the formation of the foci of magma generation. The driving forces of the process may differ however depending on the mechanism of rifting. In active rifting the formation of the foci of magma generation is connected with the local thermal anomalies in the asthenoliths roofs. In passive rifting the prevailing value, obviously, has the decompression of the upper mantle rocks, which facilitates the development of the processes of the partial melting. Thus, the key factors of the evolution of magmas at the focus stage are temperature, pressure, quantity and composition of fluid, and also the degree of the heterogeneity of the substratum of magma formation. In different CDZ these factors can be characterised by different-scale variations, causing the specific character of magmas already at the focus stage. However, the reconstruction of the specific combinations of the basic parameters of magma formation is a rather complex problem. Moreover, the process of magma generation is not fully studied yet; the existing developments still include some assumptions and controversial points.
2.2. TRANSIT STAGE The beginning of the transit stage of the pre-eruptive magma evolution can be, obviously, connected with the moment of their detachment from the primordial melting zone. The existing works show that this process can occur over a wide range of the fluid-dynamic conditions, thanks to which both the gradual and sharp passages are possible between the focus and transit stages. The most common model of the detachment of primary magmas assumes the gradual accumulation of partial melts in the apical part of the fusion zone and then the rapid ascent of the segregated melts [52, 56, 59, 69, etc.]. It is usually considered that the accumulation of melt is caused by its directed flow in the permeable medium, such as do appear the partially molten rocks, but some researchers do assume that this process is achieved in the wave regime, in the form of so-called "magmons" [60]. Some models of the detachment of magmas assume the incomplete melt segregation from refractory phases [22, 27]. Thus, in the model of poly-baric assimilation is allowed the possibility of certain quantity of residuum crystals capture from the partial melting zone by released magmas and further ascent of such solid phases in the magma suspension [27, 37]. From the other hand, the decompression-dissipation model of magma formation assumes squeezing from fusion zone of the melt and residuum crystals in the form of some pasty mixture or migma [22]. Thus, depending on the mechanism of the detachment of primary magmas two basic tendencies of their subsequent evolution can be drawn: 4
• •
the evolution of homogeneous magmas the evolution of heterogeneous melt - solid-phase mixtures.
To this one should add that in both cases it is possible the participation of the fluid component, which is contained both in the form dissolved in the melt, and, probably, as the free phase. Since during the transit stage the magmatic melt is moved up to the significant distances, it is necessary to describe the basic features of the dynamics of this process, with which the factors of the magma evolution can be functionally connected. Numerous studies of the physical aspects of magmatic process show that the dynamic behavior of the magma transport is determined in many respects by the forces, which generally cause any melt uplift in the lithosphere. However, nature of these driving forces remains the subject of discussions yet. Commonly for their explanation is drawn the model of the gravitational instability of the melt of the zones of the partial melting relatively to the heavy residuum component. It is considered that as a result of creating the significant difference in the densities the segregated partial melts (usually in this case the basaltic melts mean) they begin to float up in the denser surrounding peridotite rocks. Exactly the same mechanism and its driving forces also is commonly used for the simulation of the mantle diapirs ascent, that are nascent in the deep levels of upper mantle and on its boundary with the lower mantle. The stage of the magma detachment in this case corresponds to the moment of achievement by the adiabatically ascending asthenolith of a certain critical level. It is higher than one when the melts accumulated in its apical region part become gravitationally unstable with respect to not only surrounding rocks, but to the asthenolith itself. In spite of the apparent consistency of this model, it is not deprived of some sufficient imperfections. In particular, the simulation of the magma ascent under the effect of the density inversion is conducted mainly on the basis of the parameters that very approximately describe the real system. Moreover, it looks likely that the role of the density inversion as the cause for the vertical transport of melts is exaggerated inadequately. At the same time, other important circumstance mainly does not get proper attention. The question is that majority of silicate liquids are relatively poor in volatile components. Hence in partial melting the liquids undergo not only density decrease but also volume increase, according to the Clausius-Clapeyron law dT/dP = dV/dS. The estimation of the magnitude of the volume effect of the reactions of silicate melting is a complex problem that is essentially caused by uncertainty with respect to the quantity and composition of the volatile components, and according to some data a volume gain can reach 15-20% [13]. It is obvious that the creation of excess volume in the zones of magma generation can render as perceptible effect on the primary magma ascent as its gravitational instability does. Moreover, the volume effect of the reactions of melting is capable, probable, in a most direct manner to act on entire course of the process of magma generation. It was V.Khlestov [18, 19] who had paid special attention to this circumstance and have showed that on the basis of the mentioned law of Clausius-Clapeyron, the temperature increment or pressure, which facilitates melting, must be compensated by appropriate increase of volume. This condition for the case of continuous medium, corresponding in basic parameters to the upper mantle, is reached probably not always. Therefore the inability of the volume effect release of the melting reactions can considerably hold in control the processes of melting, displacing the solidus line of substratum by 200-300oC into the high-temperature region, which must lead to the appropriate overheating of primary magmas relative to their liquidus curve. Since such considerably overheated magmas are rather rare [1, 18], it is possible to expect that during the magma formation are nevertheless created the conditions for the realisation of the volume effect of the melting reactions by formation of the necessary excess volume in the rocks hosting the melting zone. This result can be reached both due to the internal dynamic potential of the melting zone itself and due to the external with respect to it (tectonic) factors.
5
The volume effect of the melting reactions causes the appearance of the excess pressure of melt on of the surrounding rocks. This pressure in conjunction with the pressure, created by a difference in the densities between the melt and the surrounding rocks, and with the pressure of the fluid dissolved in the melt, form the Internal Dynamic Potential (hereafter IDP) of the melting zone. In certain cases IDP can, apparently, exceed the lithostatic pressure of the overlapping rocks and cause raising roof rocks of focus without the essential disturbance of their continuity. It can be supposed also that uplift of the mantle diapirs in the active rifting is caused by the same effect. Each separately undertaken asthenolith can be associated with the magma chamber, which has very large dimensions, and which contains comparatively small amount of melt (~10-15%) and dissolved fluid. The combination of these driving factors provides the IDP, which is sufficient for a certain rising of the overlaying sections of lithosphere. However, as far as strictly the foci of magma formation, which complicate the growing asthenoliths, are concerned, their IDP, apparently, is not sufficient for raising the blocks of lithosphere. More probable that this IDP can put the local stresses causing the destruction of the overlapping rocks and the penetration of magmas into weakened zones resultant in this case. In other words, the magma uplift from the foci of magma generation is accomplished, in the opinion of many researchers, as a result of the melt invasion into the lithosphere crushing zones that have been formed due to these melt strong impact. The latter process is described well by the model of hydraulic rupture or, strictly speaking, of magma rupture [41, 60-63, 65]. The ground of this phenomenon provides the Rehbinder effect, which consists in lightening of the solid bodies deformation and destruction, including rocks, as a result of reduction in the free surface energy on the contact with the liquid phases (see overview in [15]). In works [61-63] it was shown that magma rupture is controlled by such critical parameters as viscosity, density, and especially the pressure of liquid or mixture of melt and fluid, that move on the weakened zone, which usually is associated with the crack. These characteristics determine the rate of deformation of the overlapping rocks and respectively rate of the melt propagation. The analytical idea of the existing interconnections between the values indicated is given in works [41, 61-63, 65], so here it is worthy to note that rate of propagation is maximum for the rich in volatile components kimberlite, lamproite and related magmas. Somewhat less rate have the alkali-basalt magmas. According to the estimations of F.Spera [65], in the latter case rate of magma ascent composes value on the order of 0.1 m/s whereas for the kimberlite magma the rate can reach 5 m/s. It looks likely that the rate of tholeiite magma ascent can be close to the one of the alkaline basalts, but smaller fluid content of tholeiites gives some grounds to assume that their uplift occurs with the lower speed (~0.05 m/s). In contrast to this significant fluid content of the andesitebasaltic and andesite magmas indicates high rates of climb that, in particular, frequently occurs during the eruptions of these magmas (confined, however, to the construction (collisional) geodynamic situations like island arcs). However, as far as more acid melts are concerned, in spite of the potential possibility of the content of significant quantities of volatile components, their uplift is apparently hampered by high viscosity. Therefore for the granite magmas, which are usually represent the melt - solid-phase mixtures [22], it is assumed the uplift predominantly in the form of diapirs [22, 53], for which the rates of growth, probably, by several orders less than such of more mafic magmas. At the same time, some portions of the felsic melts, which are characterised by homogeneity, are obviously capable to rise at velocities, commensurate with the same for the basaltic magmas. In this case the extrusion is not caused by IDP of the corresponding melting zones, but by some external forces (in more detail this question it will be discussed in the examination of chamber stage). Although the developed models of the magma ascent forced by hydraulic rupture convincingly describe the mechanism of the IDP realisation, it is obvious, that the mentioned external factors also can provide significant effect. For example, creation of the compressive stresses field around the melting zone can considerably increase the pressure of the melts, which are characterised by low compressibility, and thus contribute to their intrusion into the overlaying rocks by the mechanism of magma rupture. However, this 6
situation is inherent in the larger measure in the collision geodynamic situations. In CDZ, the general tension of lithosphere does cause, from one side, the decompression of melting zones, which causes additional magma generation [5, 22]. And from another hand it leads to reduction in the strength of the rocks, which overlap melting zones, and facilitates the process of magma rupture and magma ascent. A question about the melting zones IDP and external tectonic factors leaves far beyond the framework of the dynamics of the magma uplift. It is evident from above that between IDP and external tectonic forces two principal relationships can be expected, whose specific pattern has cardinal value for geodynamics as a whole. But in the majority of modern geodynamic models it is assumed the primacy of the actively rising deep substance under the effect of the internal energy, either it be the convective process as a whole or its special case of diapirism. The tectonic motions of local and regional nature are considered as derivatives. Usually from the passive rifting model it follows the primacy of the regional tension of lithosphere, which leads to the asthenosphere uplift in the region of the greatest decompression of the upper mantle. Consequently, in certain cases of the deep substance uplift it is possible to assume the primacy of the tectonic factors that directly displayed, for instance, in deep-seated faults as trans-lithospheric channels of the transport of mantle magmas, as this was noted by some researchers [16, 17]. Nevertheless it should be noted that in this case nature of regional tectonic motions is not specified commonly. Meanwhile these motions can be the derivatives of the significant volumes of substance displacement in the deepest parts of the upper mantle or even in the lower mantle (Fig. 1). Fig. 1. Possible relationships between the active (a) and passive (b) rifting in case of common asthenolith origin but different levels of the asthenolith outward spreading Indicated coloured areas: in green - Earth crust; in magenta - upper mantle; in grey - lower mantle; blue mushrooms - mantle diapirs. Yellow arrows - primary convective motions in deep mantle regions. Vertical movements inside mantle diapirs shown with cyan arrows. Induced tangential movements shown with red arrows.
Hence it is apparent that even passive rifting allows the primacy of the deep substance uplift in the form of convective flows or diapirs. Thus, the basic reason for the mantle magma ascent in the lithosphere is the centrifugal movement of deep masses, which is achieved in the form of the convection currents or diapirs. In this process the mantle melting zones provide additional IDP of the ascending magmas, which facilitates their intrusion into the overlaying rocks. This process is controlled, however, not only by IDP and by the rheological features of magmas, but also by the mechanical properties of the surrounding rocks of lithosphere. The summary effect of lithostatic pressure and increase in the temperature with the depth causes three main conditions of deformation in the lithosphere: plastic, semi-brittle and brittle. For the upper part of the lithosphere are characteristic the brittle deformations, which in proportion to sinking are changed by semibrittle (micro-fracturing in combination with the elements of flow) and further by plastic ones. The regime of semi-brittle deformations is, thus, transitional between one controlled by the stresses (brittle deformations) and one controlled by the thermal conditions (plastic deformations).
7
Fig. 2 depicts the diagrams of the distribution of the types of deformations in the lithosphere for the situations of tension and compression. It is evident that the power of the zone of brittle deformations is very significant for the situations of tension. In CDZ this will cover almost entire sialic part of the lithosphere. The zone of semi-brittle deformations also has enough thickness, which in the case of CDZ can be associated with the upper parts of upper mantle. In contrast to this in the situation of compression the zone of brittle deformations covers only the near-surface levels of crust, whereas its large part is found in the semi-brittle state. Fig. 2. Stress distribution in the lithosphere of the extension and compression geodynamic environments Simplified after S.Kirby, 1980 [ 50]
Coloured areas show different types of deformation
It should be noted that the mechanical properties of the rocks of lithosphere can, apparently, vary under the effect of IDP of magmas, which ensues, in particular, from the model of hydraulic or magma rupture. Obviously, the high values of the magma pressure can cause reduction in the strength of the surrounding rocks to the linear expansion and convert them from the more plastic to more brittle state. In other words, in the areas of the magma uplift, which possess high IDP (for example, kimberlite), the zones of the brittle and semi-brittle deformations of the rocks of lithosphere can be extended considerably deeper than this is shown in Fig. 2. Thus, in the destructive zones the magma uplift is considerably facilitated by the domination of brittle and semi-brittle deformations over the large thickness of lithosphere. Idea about the morphology of the magma transport channels appearing under these conditions can be composed in the examination of the system of the tectonic breakdowns, which are formed in the crest part of the idealised diapir. Fig. 3. Simplified sketch of the classic rift zone development in the crestal region of the rising mantle diapir (left) and resulted stress distribution in the lithosphere (right) Left: Earth crust shown in light green, mantle diapir in blue. Right: Stress release with respect to: σ1 - maximum stress axis; σ3 - minimum stress axis. Tension fractures shown in red, shear fractures indicated in bright green dashes.
Fig. 3 shows the distribution of the maximum σ1 and minimum σ3 stress axes in the process of the idealised rift zone development over in the crest part of the rising mantle diapir. It is evident that orientation of the main fracture network to the larger degree corresponds to 8
the arrangement of the planes of the shear cracks in the strain ellipsoid. It is known that the shear fractures are characterised by tight limb closure and the shear zones are characterised by a comparatively small permeability. The tension fractures are considerably more permeable, but their magnitude is rather inconstant and it depends greatly on the anisotropy of the medium. Therefore it is possible to expect that in the actual conditions the most favourable situation for magma uplift is created in the sections of the development of the contiguous system of the tension fractures, where the cross connections between them make possible due to magma rupture. Obviously, the main channels of the magma transport form in this way in the zones of deep-seated faults. However, the relatively free magma uplift along the tension fractures is limited in the sections of the domination of the shear fractures, and commonly the movement of the melt is achieved in the regime of magma rupture along the planes of the shear fractures. Thus, most important factors, which control the regime of the magma uplift are their IDP, rheological properties, and also mechanical state of the surrounding rocks of lithosphere. Each of these parameters can vary in the sufficiently wide limits even for a single geodynamic situation exemplified by the CDZ. In turn, such variations can render different effect on the evolution of magmatic melts. It is obvious that the homogeneous magma, which was isolated from the melting zone, possesses its own enthalpy and is deprived of the additional energy sources. This determines the basic trend in the evolution of homogeneous magmas, which is evinced by gradual reduction in the enthalpy. This process is achieved due to the heat expenditures for adiabatic expansion and loss into the enclosing rocks. It was shown [5, 22] that the adiabatic expansion is accompanied by decompression (the most probable for the destructive zones process) and it is usually small in term of the heat losses. As a result the magma can keep sufficiently high temperatures at the high crustal levels without the essential crystallisation. Heat emission into the enclosing rocks is one of the most effective factors of the magma cooling and crystallisation [20, 21, 40, etc.]. These studies have shown that the critical parameters of the process of magma heat emission and cooling are the rate of flow and the rheology of magmas, thermal conductivity and melting point of the surrounding rocks, and also the difference of the latter and the temperature of magma. The particular combinations of the critical parameters determine two basic regimes of magma cooling during uplift: • •
partial crystallisation with the retention of contact with the surrounding rocks extensive crystallisation in the boundary parts and on the walls of the channel.
In the first case the retention of the contact of magma and wall rocks has high value not only for the heat transfer, but also for the mass transfer, i.e., the contamination of magma by the material of the surrounding rocks. In CDZ this factor acquires special role in the modification of the mantle magma composition because the differences in temperatures of mantle magma and the solidus temperatures of the rocks of sialic crust are sufficiently great in order to ensure noticeable assimilation of crustal material by the mantle magma. I.Campbell has showed [33], that the thermal erosion of the walls of channels in the continental crust by the rising mantle magmas is most effective in the case of the turbulent flow of the melts. The width of channel and the specific rheological parameters of magmas are the condition of this flow pattern. According to the calculations of this author, for the primary basaltic magmas with the width of channel of about 3 m, the viscosity of 10-102 poise and density 2.7 g/cm3 it is possible extensive thermal erosion and assimilation of the material of the walls of dioritic composition. It is characteristic that in this case the powerful convective heat exchange causes the temperature balance of the basic volume of magma and its insignificant crystallisation. Simultaneously turbulent flow conditions prevent the segregation of solid phases and melt. Thus, the regime of magma cooling with the retention of contact with the surrounding rocks 9
is characteristic, apparently, for the magmas, formed in the foci with high IDP. They are characterised by such relationships of the most important rheological parameters (viscosity and density) which are sufficient for guaranteeing the turbulent flow. Furthermore, these magmas are connected with the destructive geodynamic situations, where, as it was shown, optimum conditions for the development of significant brittle deformations are created (channels of the magma transport). The evolution of magmas under the effect of the fractional crystallisation in this case, apparently, is insignificant. The noticeable modifications of the composition of deep melts occur in CDZ as a result of the assimilation of sialic crustal material. From the above it follows that the regime of cooling with the intensive crystallisation in the marginal regions and on the walls of channels can be achieved only in the sufficiently long period. In this case is established more or less stationary temperature gradient from the center of channel to the walls and further into the enclosing rocks. Probably, this regime is characteristic for the magmas with the high viscosity, which are formed in the foci with relatively small IDP and confined to the convergent (collision) geodynamic situations. At the same time this cooling regime can be established in the channel after the prolonged stage of the magma flow, for example, after the exhaustion IDP of the melting zones. In other words, during a single volcanic cycle one regime can replace another, the second of them corresponds, apparently, to the interruptions in the volcanic activity. Cooling with the crystallisation in the boundary parts and on the walls of channel leads to the extensive fractionation of the melt. The solid phases segregation can be achieved both due to the sinking of crystals and as a result of their growth on the walls of channel. At the same time the assimilation of crustal material in this case is minimum, since the melt is isolated from the surrounding rocks by the crystalline framework of cotectic minerals. Besides the described thermal factors, one should dwell on another important factor of the magma evolution, that is regime of volatile components, and water first of all. The fluid dissolved in the magma, as is known, substantially reduces the solidus temperature of the magma. The degree of reduction is directly proportional to a quantity of fluid [1, 5, etc.]. Since the solubility of fluid strongly depends on pressure, at the particular depth it become possible the fluid separation into the free phase and the boiling of melt, which is especially characteristic for strongly wetted magmas [5, 22]. In this case sharp reduction in the solubility of fluid does throw the magma in the direction its dry solidus. This causes the rapid crystallisation of melt in the sub-isothermal conditions. It should be noted that known examples of the combined eruption of the lavas and fluid phase, and also the diverse pyroclastic and explosive formations indicate that the fluid separation into free phase is not always accompanied by its complete detachment from the primordial melt. In looks likely that after the boiling of melt and its some rapid crystallisation further ascending motion of the system in the form of three-phase melt-fluid-crystalline suspension occurs. Free fluid phase in this case is capable, apparently, of ensuring the retention the viscosity of the system at the level of primordial melt, and possibly even, in the case of the supersaturated in fluid magmas (alkaline series), to reduce the general viscosity. Therefore after melt boiling the regime of its uplift can be inherited from its earlier one, which match two types examined above. From this point of view in the highly dynamic regime of uplift it is possible to expect the larger degree of crystallisation that, however, does not indicate the segregation of the crystals, which can be carried in the indicated suspension as intra-telluric phenocrysts. However, in the low-dynamic regime of uplift after the melt boiling it is possible to expect rapid crystallisation both near the walls of channel and in the center section, up to the total termination of the crystallised melt uplift. In the beginning of this section it was noted that in the modern models is examined the possibility of detachment from the melting regions not only of the homogeneous, but also heterogeneous magmas, which are exactly melt - solid-phase mixtures. In contrast to the homogeneous magmas the motion of such mixtures is accompanied by the dissipation of the heat of the friction of viscous flow [22]. This additional energy source in combination with the adiabatic decompression can provide secondary melting the solid phases of the mixture up to its complete homogenisation, as this is assumed, in particular, with the 10
decompression-dissipation model of magma formation [22]. After the homogenisation of magma its evolution occurs according to the systems described above. Consequently, the evolution of heterogeneous mixtures is characterised by the additional stage of relict solid phases melting beyond the limits of the magma generation regions. It is obvious that these phases cannot be always melted completely and in the following stage of the magma cooling and fractionation they can play the role of the embryos of the crystallisation of comparatively low-baric minerals. The extension of relict phases into the high crustal horizons is possible, probably, only in the highly dynamic regime of magma ascent. Thus, in the process of uplift from the melting regions the magmatic melts can undergo different kind of modifications. The evolution of magmas at the transit stage is controlled by such most important factors as IDP of the melting zones, rheological properties of the melts, regime of volatile components, and also the mechanical properties of the rocks of lithosphere. CDZ are characterised by the high permeability of lithosphere, ensure the maximum realisation of the melting zones IDP. Therefore it is possible to expect that the magmas of CDZ rise in the highly dynamic regime, unfavourable for the significant fractional crystallisation, but which facilitates the assimilation of sialic crustal material. At the same time the alkaline series magmas rich in fluid components in the process of uplift can experience significant crystallisation as a result of the of melt boiling.
2.3. CHAMBER STAGE Idea about the transitional magma chambers is classical in petrology and finds confirmation with the aid of both the geophysical methods and petrological and geodynamic reconstructions. Transitional chambers represent some reservoirs of the magma conservation. From the discussion in previous section, the appearance of such reservoirs in CDZ seems to be rather improbable for the first look. Nevertheless, according to calculations of J.Crisp [38], for the rift zones the ratio of the volumes of extrusive and intrusive magmatism has a value about 1:4. Therefore before going on to the description of the special features of the magma evolution in the transitional chambers, it is necessary to discuss briefly the conditions of such reservoirs formation in CDZ. The basic problem of the magma chamber formation pointed by R.Daly [39] for the first time is a question of the excess volume, filled with magma. Together with the concept of roof caving by R.Daly, are now examined also the models of igneous diapirism and zone melting [53], applied to the zones of plastic deformations, and the roof uplift [70] in the region of brittle and semi-brittle deformations. Based on the general mechanical situation in the lithosphere of CDZ, described in the foregoing section, the last model appear to have been taken into account first of all. The roof uplift is possible when the magma pressure exceeds the lithostatic pressure of the overlaying rocks. In this situation one can expect the uplift of the separate blocks of lithosphere, limited by the planes of the shear fracture system. In connection with CDZ the possible scenario is depicted in Fig. 3, and the sketch of the chamber formation by this model is given in Fig. 4. Fig. 4. Scheme of the magma chamber formation upon the introduction of the magma, which possesses the significant value of pressure (IDP), with uplift of the overlaying block a - the common shape of magma chambers according to [70] b - the sketch of chamber formation. See Fig. 3 for orientation of the shear fractures (shown in red), which define the block detachment and magma movement in this case.
11
As is evident, the shape of magma chamber, which is considered as the most common (Fig. 4a), can be very simply obtained upon the intrusion of magma under the high pressure with the uplift and the partial caving of the roof, which stops the chamber up. It follows from this scheme that the higher is the magma pressure, the deeper is the level, at which is possible the plumbing block uplift. Consequently, high-dynamic kimberlite and alkaline-basaltic magmas can, probably, form transitional chambers at the considerable depths (~30-50 km), whereas for the tholeiitic melts are more typical the shallow-level transitional chambers. It should be noted that in addition to the aforesaid the transitional chambers can be subdivided into those being periodically refilled and sealed varieties. Second type of chambers obviously corresponds to the so-called stratified intrusions. The formation of such chambers can be, apparently, connected with the low-dynamic regime of the magma ascent. In contrast to this the periodically refilled magma chambers are the constituent part of the trans-lithospheric magma columns, which feed surface volcanism, and are connected with the high-dynamic regime of the magma uplift. For the long time the processes in the periodically refilled chambers were reconstructed on the basis of the experience from the stratified intrusions. The magma evolution in this case was described by the model of the mineral gravitational accumulation in the bottom portion of the chamber. However, a question remains unresolved as to the accuracy of the application of a model of the closed system (stratified intrusion) to the open system, i.e., refilled chamber. Furthermore, it was required to be understood a special mechanism of the separation of light crystals (plagioclase) and their bottom accumulation in the layer of denser magma. In the last decade the English, Australian and American petrologists have carried out a number of experiments, which imitate processes in the periodically refilled magma chambers. It was proposed a set of the evolutionary schemes, which make possible not only to compose the more substantiated idea about such chambers, but also in many cases to re-interpret the mechanism of the classic stratified intrusions formation. The principal distinction of the new models is the idea about fractionation between not solid and liquid phases, but between the liquids of different composition [54], i.e., primary magmas and their derivatives. During fractionation of some primary melt in the chamber it is formed certain quantity of the residual liquid, that differs in physical properties from the primary magma. Such differences can be of two kinds that are determined by the specific path, which the magma follows - Bowen or Fenner ones. Fig. 5 gives the expression of a change in viscosity and density of the different magmas in the process of differentiation. These diagrams show that the density of magmas is the most extensively changing parameter. The crystallisation of primary magmas of calc-alkaline and alkaline series, which follows the Bowen path, is accompanied by the formation of less dense residual melts. During the crystallisation of the tholeiite primary magmas according to the Fenner path less dense residual melts are being formed at the early and last stages, when among the crystallising solid phases a great proportion occurs of dark-colored minerals, in this case of olivine and magnetite respectively. These two stages are separated by the section, in which the density of residual melts grows up due to the involvement of significant quantities of plagioclase in the crystallisation process. Thus, once the portion of primary magma with a density r0 enters the transitional chamber, the density difference between new magma and chamber residual liquid (ρ1) may attain as positive (ρ0- ρ1 > 0) as, probably rarely, negative (ρ0- ρ1 < 0) value.
12
Fig. 5. Some rheological parameters of the magmas of the most common types.
a - change in the melt viscosity in proportion to the temperature increase after [42]; The melts: R - rhyolite, A andesite, T - tholeiite, AB - alkaline basalt. b - change in the density of the melts of basic magmatic series depending on the fractionation degree after [26]. C - calc-alkaline series, A - alkaline series, T - tholeiitic series.
In the first case the chamber become a two-layer system, which includes the upper layer of the lighter, more viscous cooled differentiated melt and its underlying layer of hot, denser and less viscous primary magma. Joint cooling of both layers begins after the stabilisation of the chamber. The most extensive heat emission is achieved, obviously, through the roof parts of the chamber. Therefore it would be possible to expect that the crystallisation begin in the upper melt, as this is assumed in the classical accumulation model. However, upper melt, as has already been spoken, is more fractionated, and its cooling degree seemingly is buffered by lower liquidus temperature. Furthermore, one must take into account that the lower melt must be cooled via heat transfer into the upper layer (Fig. 6). Fig. 6. Zonation in two-layer chamber and the distribution of convective flows in each layer
Simplified after [26].
Consequently, the lower melt layer will stay sufficiently long time in the hyper-liquidus, possibly, even strongly overheated state. In this case it should be noted that the extensive heat transfer is accomplished without any essential mass transfer between two layers, i.e., the rates of diffusion of the heat and the matter are different. This interface was called the double-diffusive one [36, 47]. In both layers in this case there are intensive convective flows, which ensure vertical heat transfer. Since these flows occur autonomously, each of the layers can be considered as some convective cell, which also was called double-diffusive. Thus, prior to the beginning of the post-injection crystallisation the chamber will include two cells of the double-diffusive convection, separated by double-diffusive boundary.
13
At some moment of time the temperature of lower melt will be lowered to the liquidus. As a result of extensive vertical heat transfer this point is reached in the basal part of the lower layer, where begins in situ crystallisation [32, 55]. The initial stages of the crystallisation of primary magmas both over Bowen or Fenner path it is accompanied by the formation of the lighter residual melt (see Fig. 5). The formation of this melt in the bottom part of the lower layer of primary magma will cause the gravitational destabilisation of the latter and the convective floating up of the differentiated melt. Experiments and calculations [42-44] show that this floating up is accomplished in the turbulent regime, thanks to which small crystals will be carried by convective jets from the zone of crystallisation, and the fresh, undifferentiated melt will input there instead. The ascending convective flows of the fractionated liquid, being mixed up with the primary melt, which did not reach the crystallisation zone, produce the hybrid melt of transitional density. This melt will be located in the upper part of the lower layer, forming the independent micro-cell of double-diffusive convection. The suspension of crystals with time is carried into the bottom part of the chamber by convective downstreams. While the crystallisation in the bottom part of the chamber proceeds, in the main volume will occur progressive replacement of the initial magma by the fractionated melt. Furthermore, it will produce increasing stratification of the lower part of the chamber into the large number of micro-cells of the double-diffusive convection. In each cell the density of the melt will decrease upward. Crystallisation and double-diffusive convection in the lower layer will continue until the composition of the liquid of the uppermost micro-cell of the double-diffusive convection of lower layer becomes equal to the composition of the differentiated magma of upper layer. At this moment will occur mixing the both melts and formation of the united cell of the double-diffusive convection. A similar combining of the earlier and the newly formed differentiates will occur until relative homogenous state of the melt is achieved in the chamber, i.e., until single cell of the double-diffusive convection of the differentiated melt is formed. Thus, in the case of injection into the chamber of the portion of less dense primary magma the composition of the differentiated melt of the upper layer is supported to be kept approximately on the same level. The upper layer itself is constantly grows up from below due to the addition of the micro-cells of the double-diffusive convection. In the second cases of injecting the primary magma under discussion, when the density difference ρ0-ρ1 has negative value, the system after injection immediately becomes gravitationally unstable and primary magma begins convectively float up within the denser differentiated melt. Mixing of both melts in this case is possible if the turbulent regime of floating is achieved [45], whereas during the stream-line conditions primary melt floats up in the form of uniform column without the mixing. The differentiation of residual melt at the stage of the formation of denser derivatives, characteristic for the tholeiitic series, also goes on via the cells of the double-diffusive convection, but in a somewhat different form. The residual melts of the increased density will be concentrated directly higher than the level of the formation of cumulative phases, but without the mixing with the melt of the main volume of magma. In this case the growth of the cells of the double-diffusive convection will be achieved from below, whereas in the initial stages of fractionation - from the top. The evolution of the melts in the periodically refilled magma chambers is determined not only by the changes in the compositions of melts, but also by dynamic patterns of their injections, or by shape and size of the reservoirs. The feeding dynamics of the magma chambers is studied experimentally and is described theoretically [34, 35, 48, 70]. Like any other cases of liquid flow, this process can be achieved in the laminar or turbulent regimes. Reynolds number [35] provides the criterion of flow conditions:
14
Re1 = wd/kv1 , where w - the average speed of flow in feeder (factors, which control this parameter, were examined in the previous section), d - diameter of the channel, v1 - kinematic viscosity of the fresh portion of magma, k - proportionality factor. With the values of the number of Re1 < 30 inflow of magma is achieved in the completely laminar regime, with Re1 > 400 - in the completely turbulent one. The transitional values of the number Re1 define the zone of magma flow with the elements of both the laminar and the turbulent of regimes. Spreading of the melt in the bottom part of the reservoir occurs with the laminar inflow of the new portion of magma into the chamber. The specific level of the primary magma spreading depends on the degree of the differentiation of the melt of lower layer in the chamber up to the moment of injection. If injection occurs under ρ0 > ρ1, i.e. before the considerable growth of the density of residual liquid after the passage of its density minimum (see Fig. 5), the dense injected melt will be arranged in the lowermost part of the liquid volume of the chamber. The chamber itself in this case will become three-layered. In the lower layer will be located the fresh portion of magma, in the mean level - partially fractionated melt with a series of the cells of the double-diffusive convection, and upper layer will be preserved constant, inherited from the two-layered chamber. But if injection occurs under ρ0 < ρ1 into the lower strongly differentiated layer of two-layered chamber, then primary melt, as has already been spoken, will convect upward and it will be arranged directly higher than cell of the double-diffusive convection of the melt of the maximum density, when is achieved the condition of ρ0 > ρ1, [45]. One should emphasise that in both cases under laminar conditions of injection the mixing of fresh and differentiated melts is absent or very insignificant. The injection of fresh melt in this regime leads to the formation of the additional layer in the chamber, in which since the beginning of the fractionation appears a series of the cells of the double-diffusive convection. Mixing of the melts in the chamber occurs as a result of the combined fractional crystallisation in each of the layers. Upon the entering of the new portion of magma in the turbulent regime injection itself is accomplished in the form of the fountain of varying height [35]. This nature of introduction assumes the possibility of active interaction of the injected magma and residual melt in the chamber. The criterion, which determines the possibility of turbulent mixing of both melts, takes the form of modified Reynolds number [35]: Re2 = wd/v2 , where w - the speed of magma flow within the limits of fountain, d - the average diameter of the fountain, v2 - kinematic viscosity of the host melt in the chamber. With the values of Re2 < 7 the mixing of the melts is absent or very insignificant, and the turbulent flow of magma is accomplished in the limits of fountain, without affecting the surrounding melt. After braking of the injected melt it descends into the bottom part of the chamber, where it is located at the level, dictated by the distribution of the melt densities in the lower layer. Extensive mixing of both melts will occur with the values of Re2 > 70 [35]. The transitional values of 7 < Re2 < 70 characterise the region of the increasing mixing of melts. Thus, criterial values are distinguished between themselves by an order. This means that at the lowest possible rates of flow of magma in the fountain the unlimited mixing with the host melt is possible if and only if the viscosity of the latter differs from the viscosity of the injected magma not more than by an order. This is why I.Campbell and S.Turner [35] drew the conclusion that upon the turbulent introduction of primitive basaltic magma into the chamber, filled to a considerable degree with the differentiated basaltic melt, mixing of both melts is possible on wide scales, whereas the introduction of the same melt into the upper viscous, for example, rhyolitic, layer of chamber, will not lead to the mixing of melts. Thus, the dynamic regime of the injections of the fresh portions of magma controls the degree of the breakdown of the conditions of fractionation in the stabilised chamber. In other words, it determines possibility and scales of the mixing of the primary and differentiated magmas with the formation of the hybrid melts, which are located in 15
accordance with the distribution of the densities of the melt of the lower part of the chamber. The nature of the evolution of magma melts in the chambers in many respects depends also on shape and size of the latter. Dependence on the chamber shape is caused by different direction of motion and fractionation of the melts in the boundary parts of the reservoirs near the vertical or inclined walls. Usually three versions are being examined as to the shape of the chambers (Fig. 7), between which are possible the rather diverse combinations. Fig. 7. Magma chamber wall shapes
a - chamber with the vertical walls; b - chamber with inward inclined walls (glass-shape bottom); c - chamber with upward going walls (gabled-shape roof).
The magma evolution near the vertical wall is studied in very details [26, 54, 57, 71]. As in the previously models examined, fluid dynamics is determined mainly by the relationships between the densities of the magma in the main volume of the chamber (ρk*) and melt, which is formed during the crystallisation of this magma of (ρm). With the appearance of a difference in these densities near the wall is formed the boundary layer of low thickness. According to the estimations of R.Nilson et al. [57], at the height of chamber wall in 1 km the thickness of boundary layer composes just a few centimeters. With the positive value of a difference in the densities of ρk - ρm > 0 the differentiated melt will float up in the limits of boundary layer along the wall to the upper part of the chamber, where, being stored, it will form independent layer with a series of the cells of the doublediffusive convection (Fig. 8). Fig. 8. Double-diffusive convection in the magma chamber with the vertical walls (adapted after [ 57]) 1. host rocks 2. convection currents in the main volume of lower layer 3. conditional axis of zero speeds and the speed distribution profile of convection currents in the main volume and in the boundary layer 4. main double-diffusive interface 5. double-diffusive interfaces in the upper layer 6. ascending convective jets in the boundary layer
In this case the motion of the differentiated melt will create reverse flow with respect to the branches of the double-diffusive convection in the main volume of chamber. The rate of flow in the cells of the double-diffusive convection is higher in general than the ascent velocity in 16
the boundary layer, moreover for the latter the maximum values of speeds are observed near the margin of the boundary layer and main volume of the magma. Besides fractionation of liquids due to the differently directed flow, in the case under question the separation of solid and liquid phases can be effective as well. This is because descending branch of the double-diffusive convection can capture some of the sinking crystals, formed near the wall. In this way the value of these crystals may increase in the region of the most extensive crystallisation in the bottom of the chamber. With the negative value of a difference in the densities ρk - ρm < 0 the differentiated melt will descend to the bottom of the chamber, where it can form basal layer and undergo further crystallisation. Thus, crystallisation near the vertical wall does occur without a noticeable change of the composition of the melt in main volume of a chamber and can cause concentration in its roof of the certain quantity of a comparatively light differentiate, and in the bottom - of dense melt and the crystals, transferred by the convection currents of the main volume of chamber from the place of their formation at the walls. The evolution of melt near the inclined chamber wall is characterised by a number of distinct features [46, 51]. In this case it is necessary to distinguish two cases of the contact of magma with the wall, that is one of magma location above the wall and vice versa (see Fig. 7). When inclined wall overlaps the magma, i.e., in the chamber with the inclined roof, the processes as a whole are analogous to the examined above for the vertical wall. In the case, when a chamber has the bottom inclined the evolution of the melt both near this wall and in the main volume of magma go on somewhat differently. This is caused by the fact that the light melt, which is formed during the crystallisation near the wall, becomes gravitationally unstable relative to the overlaying denser melt and it is forced to move upward, experiencing convective mixing with the main volume of magma (Fig. 9). Fig. 9. Schematic diagram, which illustrates the nature of circulation in the cells of the double-diffusive convection near the crystallization front at inclined wall Implication of the simulation experiments in the aqueous solution of Na2CO3 after [46]. Currents in double-diffusive cells shown in black. Red dotted arrows indicate ascent streams of the light liquid released from crystallisation at the model wall (brown slab). The scale of this mixing depends on the distance between i-point of inclined wall (bottom of chamber) and roof of chamber. Most enriched in light liquid thus has to be the roof parts of the main volume of the melt. In this case, besides the vertical gradient of composition and physical properties, caused by vertical convective flows, will appear the horizontal gradient of these parameters, caused by the directed increase in the modifying melt from the wall (bottom) of the chamber to its internal parts [46, 51]. Hence, crystallisation near the inclined bottom ensures the sufficiently rapid fractionation of the melts and formation of the complex cells of the double-diffusive convection under the action of the vertical and horizontal gradients of composition and physical properties of the melts. Besides the described processes of crystallisation, near the walls of magmatic chambers can be achieved the processes of melting the enclosing rocks, which has special importance for the crustal chambers in CDZ. Obviously, in this case will be formed the lighter melts than primary mantle magmas. The behaviour of such light melts near the walls of different slope is described above. Involvement in the process of the crustal material, not genetically connected with the primary magma, is the distinct feature of the evolution of the magma composition in this case. The formation of crustal melt near the vertical wall either under inclined roof affects only the composition of the upper layer, where this melt accumulates in 17
various or quantities [26, 44, 54]. It is expected also the significant variations in the content of trace and REE elements, caused by their selective extraction from the rocks of crust, as well as isotopic characteristics. However, melting of the host sialic rocks at the inclined bottom of magma chamber can lead to the significant contamination of the magma of the main volume of the chamber and modification of the primary geochemical and isotopic parameters of mantle magmas. Notable that net result of contamination under these conditions can be considerably more complex than in the examined earlier case of contamination in the process of the thermal erosion of the walls of channel during the magma uplift. It is because input of the crustal melts into the chamber is selective, i.e., the extraction of the separate components of sialic rocks occurs selectively and unevenly. The dimensions of magma chamber render the direct effect on the evolution of melt. The volume of the melt, which the chamber is capable to contain, is the main controlling factor. Depending on this will change the time of cooling melt to the liquidus temperature or of the temperature, necessary for the separation of the significant volumes of light derivatives. With the injection of the fresh portions of magma their influence on the modification of melt in the chamber is determined by the ratios of the volumes of the injected and residual melts and, naturally, it will be minimum with the low values of this ratio. Furthermore, the ratio of volumes indicated will determine the rate of cooling of the injected melts and respectively the dynamics of their crystallisation, formation of the micro-cells of the double-diffusive convection and the duration of their existence [42, 48]. Above it was discussed the processes, proceeding in the stabilised chambers, and also the behaviour of the melt, which is being introduced into these chambers. All the features examined have high value during the study of intrusive chambers, especially stratified intrusions, whose volume during their entire evolution remained constant or it experienced relatively small increase due to the limited number of injections of the fresh portions of magma. However, as far as the chambers, which fed surface volcanic systems and, as has already been spoken, that appear to have been periodically refilled, their evolution seems to be considerably more complex. It was determined not only by the processes of the melt fractionation and its makeup by the fresh portions of magma, but also by the removal of various amounts of liquid through extraction channel in the upper part of the chambers. It was shown in a series of special experimental studies [28, 30, 62, 66] that the dynamics of withdrawal from the chamber and the composition of the outgoing melt are determined finally by two main parameters: • •
output channel diameter pressure of the magma
In case of the two-layered chamber, which is of greatest interest, the dynamics of the melt removal briefly can be described as follows. Ones the output channel opens the melt from the upper layer go out first of all. The melt of upper layer is fixed with the dissection of extraction channel in it first of all. Simultaneously the interface between the upper and lower layers in the chamber just below the output channel begins progressively to be raised (Fig. 10) and in the specific time it reaches the output region. If the melt of upper layer only was moved away from the chamber up to this moment, then under new melts of both layers run to the output conduit, which leads to their extensive mixing and formation of hybrid melt [30, 66]. However, further ascent of the indicated interface leads to the forcing back of the melt of upper layer to the edges of chamber (see Fig. 10). The expiration of hybrid magmas, apparently, ceases at this stage and the extrusion of the melt of lower layer occurs. The dynamics of the described process depends on many parameters, which reflect both the physical properties of magmas, and geometric features of the chamber. Convenient criterial relations were obtained by F.Spera et al. [66] in a series of computer simulation experiments. In the generalised form they are given in Fig. 11. 18
Fig. 10. Dynamics of the magma withdrawal from the stratified chamber (simplified after [ 66])
a - dynamics of uplift of the interface between two layers depending on time H - height L - half of the chamber width l - half of the output channel width t - arbitrary units of time and position of interface corresponding to them. b - dynamics of changes in the composition of a melt, which is being moved away through the output channel from a stratified chamber HL - the fraction of the melt of lower layer in the hybrid mixture being moved away t - the arbitrary units of time (see a).
Fig. 11. Dependence of the magma withdrawal speed (Vm) on the value of basal normal stress (Sb)
Simplified after [66]. Blue lines show the equal values of a number A = W/w, where W - chamber width, w - output conduit width. Green lines indicate the equal values of the termination time of the upper layer melt input into the output conduit.
These results make it possible to see that with the fixed value of normal basal stress and depending on the value of the ratio A = W/w, where W - chamber width, and w - output conduit width, the dynamics of the magma withdrawal changes as follows. With an increase in a value A the vertical velocity of the magma motion decreases, duration of the extrusion of hybrid melts increases as well as and the time in which the melt of upper layer completely pushes aside from output channel and is deprived of the possibility of removal from the chamber. When talking about the squeezing out force, then the indicated stress, normal to the lower surface of magma in the chamber, can be caused by the previously examined update of the fresh portion of magma and by the creation of some kind of stamp effect. The extrusion of 19
the magma through output conduit will, thus, compensate a volume gain due to the injection with the retention of the volume constancy in a chamber. In this situation, therefore, all described processes of the magma replenishment and extrusion must occur simultaneously. As it was shown earlier, both magma update and extrusion lead in more or less extent to the breakdown magma chamber structure established during its stabilisation. It is obvious that the effect of the simultaneous action of these processes will lead to the essential modification of the chamber structure. In case of stratified pattern of the latter it obviously should cause the extensive mixing of the liquids different in composition and properties (both not fractionated yet and differentiates). The scale of mixing between the differentiated and fresh melts in this case will depend on the regime of magma chamber replenishment with fresh melts. In the laminar update the undifferentiated melt, which is confined to the chamber bottom, will reach the output conduit in considerable time, i.e., practically upon full exhaustion of the chamber. In the turbulent injection of fresh melt the formed fountain can rapidly reach high level of the lower layer in a chamber and further will be extruded through the output conduit. The injection of primary magma in the form of fountain indicates its primordial high vertical velocity. As can be seen from Fig. 11, it means the rather short time of the extrusion of the melts formed due to hybridism of the upper and lower layer melts as well as rapid elimination of the upper layer melt from the extrusion process. In other words, the magma, which enters the simultaneously exhausting chamber under the high pressure, can be rapidly involved in the process of extrusion, being avoid in this way the stage of magma differentiation in the chamber. Since the pressure of magma injections is directly related to IDP of the foci of magma generation (see Section 2.2) formation, the value of pressure and the duration of its retention at some fixed level are, apparently, the variable parameters. This means that the described process can be interrupted at any stage as a result, for example, of the IDP exhaustion of deep melting zone or an abrupt change in the tectonic situation from the tension to the compression. Every such interruption must be followed by some period of the melt gravitational stability restoration in the chamber. In this period some mixing may occur between chamber melts with the formation of a certain boundary layer (but not a boundary as before) or layers (under repeating interruptions) between the upper and lower melts [66]. Hence the magma withdrawal under the described mechanism, but from the three or ever more layers, will be progressively complicated, and resultant hybrid magmas in this case can have completely complex history of formation from the undefined number of sourceslayers. Since the beginning of the melt removal just below the output conduit in the chamber appears the ascending current, which can capture the crystals, accumulated in the bottom part, and ensure their extension in the form of phenocrysts. This possibility is determined by the value of the viscous stress, created by flow, which must exceed the weight of crystals. The capture criterion proposed by S.Blake [29], connects this dynamic factor with the distance h from the output conduit to the layer of cumulates in following expression: h = c(Qv/g'a)1/3 , where C - is a constant, Q - magma removal speed (flow speed), v - kinematic viscosity, a particle (crystal) radius, and g' is a reduced gravity, that is acceleration due to gravity g times Δρ/ρ, where ρ is a magma density and ρ+Δρ is particle density [29, p. 400]. However, according to the analysis performed by S.Blake, for the real magmatic chambers, in which the main cumulus mineral is olivine, for its entrainment and formation of olivine-porphyre lava the value of h must not exceed 20-30 m, which contradicts the actual observations. Therefore the mechanism of the redistribution of the indicated cumulus crystals can be realised in the peripheral sills, whereas the lava phenocrysts have, perhaps, another origin. In the opinion of S.Blake [29], they can be formed under injection of primary magmas at high speed, which ensures, as already mentioned, the rapid achievement of simultaneously revealed output conduits, but in this case experienced extensive crystallisation. This interpretation will agree with the results of some experiments, in which partial crystallisation of ascending convective jets were established [42, 70].
20
From the above discussion it can be concluded that the fluid dynamics of transitional chambers, which determines the principal features of magma evolution, is apparently characterised by significant complexity. It depends on a large number of variables, responsible for the magma physical properties, dynamic parameters of magma motion, and also geometric features of the reservoirs. These factors provide multi-variability of the melt evolution in the transitional chambers, which is running around two polar petrological processes - magma fractionation and magma mixing, whose comprehensive description is contained in work [70]. The differentiation of primary magmas in the chambers leads to the formation of melts outstanding in the physical properties, causing fractionation of the primary and derivative liquids. This magma splitting in the boundary portions of the chambers can ensure in a number of cases the irreversibility of process with the accumulation of derivative melt in the chamber roof. However, melt fractionation in the bottom of main magma volume of the chamber is accompanied by the convective mixing of primary magmas and derivatives during floating up of the latter. Mixing of the melts is possibly also under high-speed injections of primary magmas. Extensive hybridism can be achieved in the process of the magma removal from the stratified chambers. In this case it is rather difficult to determine the ratio of the end members with increasing of the number of interruptions in the eruptions. The stratified basalt-rhyolite chambers and large-scale hybridism of felsic and mafic melts have high value for CDZ, where the formation of significant volumes of the rhyolitic melts of upper layer is possible as a result of the anatectic mobilisation of sialic crust. Available observations suggest that practically in all regions of the development of the strongly differentiated volcanic series the percentage of the lavas, which in composition could be referred to the primary, usually it is very small. For the explanation to this distinct feature of tholeiitic series of the mid-ocean ridges E.Stolper and D.Walker [67] have assumed that the possibility of the entering of primary magmas to the surface is limited by their high density, in comparison with the terrestrial crust of even oceanic regions. In the opinion of these authors, the magma extrusion onto the surface is possible after their partial differentiation in the chamber and certain reduction in their density (see Fig. 5). Simultaneously H.Huppert and S.Sparks [43] have proposed the fluid dynamic mechanism of the density conservation of primary magmas in the stratified chambers. All subsequent studies confirmed the lawfulness of these ideas. In all probability, transitional chambers are actually the unique filtering systems, which ensure preferred entering to the surface of the lightest melts and limit the possibility of penetration to the surface of dense primary magmas. The filtering role of the chambers to a considerable extent depends on the dynamic regime of the injections of primary magmas. If their speed is high enough, the rapid magma ascent up to the output conduit and extrusion to the surface is quite possible without magma stopping in the chamber.
3. PROBLEM OF THE RECONSTRUCTION OF THE PREERUPTIVE MAGMA EVOLUTION IN THE CONTINENTAL DESTRUCTIVE ZONES It was shown in the previous sections that the magmatism of active geodynamic situations, in particular, of CDZ, are developed under the complex control of thermal, mechanical, fluid dynamic and other factors. Practically at all stages of the pre-eruptive magma evolution there is a quite real possibility of the deep melts modification. In this situation the thesis about the potential poly-genesis of any mantle magmas is legitimate. In other words, there is every reason to believe that mantle magmas once appeared in high crustal levels are formed as a result of not a single, but by combination of several processes. From the latter the main include partial melting, fractional crystallisation, hybridism and crustal contamination; last factor seems to be most important in CDZ. From the point of view of the poly-genesis of mantle magmas the development of the models of their melting (which is frequent task both of petrological and geodynamic studies) can be carried out only upon proper evaluation of the role of another factors of pre-eruptive magma evolution and elimination of the corresponding effects. This procedure assumes the 21
sequential reconstruction of the modifications of the composition of the primary magmas, which occurred at each of the stages of their pre-eruptive evolution. Of course in the overwhelming majority of the cases this task cannot be solved as a result of the insufficiency of data required or due to superposition of the different processes signs. However, these circumstances of objective nature do not mean the possibilities of failure of the correct restoration of the magma history. Therefore it is expedient to examine the tentative sequence of the reconstruction of the main factors of the pre-eruptive magma evolution in CDZ based on some examples. First of all it is necessary to note that the above discussion suggests for the different ways of the pre-eruptive magma evolution of different magmatic series and the individual complex of the control factors in each case. Therefore just establishment of the serial affinity of magmatic formations makes it possible to get the first approximation in outlining the overall sketch of the pre-eruptive magma evolution. In the evolution of the tholeiite series magmas the most important role usually is assigned to the process of partial melting in the relatively low-depth conditions and the fractional crystallisation in the shallow-level periodically replenished chambers. Since the second process leads to the significant modification of the composition of primary melts, the reconstruction of conditions and mechanism of the differentiation of magmas, which makes it possible in certain cases to recreate the compositions of primary magmas, is first priority task. It is necessary to say that the processes in the magmatic chambers described in section 2.3 were simulated mainly just with respect to the tholeiite series. The possible versions of the evolution of these melts in the periodically refilled chambers, in sufficient detail enumerated above, compose basis for the interpretation of factual material on particular CDZ (for example, see [8]). The igneous complexes, formed by the rocks of tholeiite series, are frequently characterised by the unsystematic alternation of the varieties that correspond to different degrees of the crystallisation of the assumed primary magmas. This distinct feature can be related to the different scales of the magma fractionation in the periodically refilling chamber. However, the same effect can be obtained also during the joined differentiation and the subsequent homogenisation (via mixing) of two melts that have entered the chamber in the different time [8, 43]. The study of tholeiite series in CDZ is connected with the solution of the quite important problem about the contribution of crustal material and the mechanism of interaction of tholeiitic magmas and sialic matter. As it was shown, there two main versions are possible: 1. contamination of mantle magmas during flow in the channel with the retention of contact with the surrounding rocks 2. hybridism of basaltic and felsic crustal melt during their joined withdrawal from the stratified basalt-rhyolite chambers In the first case the thermal erosion of walls will, obviously, contribute to the gross adoption of crustal material, whereas participation in the second version of crustal melts can ensure interaction of tholeiitic melts with the felsic liquids selectively enriched in a number of minor and trace elements. Furthermore, in the second case should be expected the presence of the close paragenetic (i.e. formed within single time span in a single geodynamic region) association of tholiites with the hybrid formations, which cover the spectrum of compositions from the tholeiits to the rhyolite. Some of so-called consecutively differentiated basalt-andesite-rhyolite formations satisfy this condition. Thus, just in CDZ there is a real possibility to fulfill the reconstruction of the separate elements of the tholeiitic magma evolution not only at the chamber, but also at the transit stages. Occurrence of the tholeiites, contaminated with the crustal material (i.e. the first of the examined versions), can be interpreted as the result of tholeiite magma uplift in the high-dynamic regime. Conversely, the lack of the formations of this type can be attributed to the low-dynamic regime of transit stage and the crystallisation in the boundary parts and 22
on the walls of channels. The low liquid-water content in the tholeiitic magmas cannot, apparently, ensure the boiling of these melts at the transit stage. As far as the focus stage of the tholeiite magma evolution is concerned, in this question remain many obscure moments even after the evaluation of the mentioned effects. Perhaps, principal one is the controversy of the primary nature of tholeiite magmas with the contents of MgO ~9-11 wt.% with respect to the usually adopted lherzolite mantle source under pressures more than 5-10 kbars [1, 4, 68]. Therefore in the examination of the focus stage of the tholeiite magma evolution it is necessary to account such a factor as the compositional heterogeneity of the mantle source, in particular the enclaves of eclogite lenses and veined bodies ("coronary" model [1]). This implies further need for the attraction of the decompression-dissipation model of magma formation [22] and examination of the version of the heterogeneous melt - solid-phase mixtures uplift. The evolution of the alkaline series magmas, which includes both alkaline-ultramafic and trachy-liparite varieties, obviously follows somewhat different path. Like the preceding case, first of all it is necessary to evaluate the scale of the magma modification in the transitional chambers. In this respect volcanics of the alkaline series are subdivided into two groups. First one includes rocks, which contain mantle xenoliths and, therefore, which unlikely underwent differentiation with complete separation of solid phases in the transitional chambers. Magmas of this group, obviously, were passed only focus and transit stages on their evolution. It looks likely that the latter stage has implemented in rather high-dynamic uplift regime, probably, with the retention of contact with the surrounding rocks. Taking into account the high saturation of such magmas in fluid components, it is possible to expect that the main direction of the deep magma modification is the thermal erosion and assimilation of crustal material, from one side, and crystallisation due to melt boiling at high crustal levels, from another one. The focus stage of the evolution of the magmas of this group also probably occurred with the participation of significant quantities of fluid components, provided by mantle metasomatism. The generation of primary melts was accomplished at the large depths (~150-200 km) in the region of plastic deformation of lithosphere. These conditions do limit the separation of melt solid-phase mixtures being favourable for the migration of fluid-melt mixtures. The latter at higher levels in upper mantle have captured certain quantity of material of the surrounding rocks in the form of xenoliths. Another group of the alkaline magmas was probably generated under smaller IDP, which caused the less dynamic regime of their uplift at the transit stage. Some crystallisation in the boundary parts and on the walls of channels can be expected as well. Such magmas did not possess the necessary energy for the trans-lithospheric motion by the mechanism of magma rupture, and at the specific levels, in the region of the development of brittle deformations (in sialic crust) the chambers of differentiation were formed. In their fractionation patterns such alkaline magmas are close to the tholeiitic (see Fig. 5), which makes it possible to expect the formation of the differentiated alkaline series in the periodically refilled chambers according to schemes examined above. In this case a special problem concern kimberlite, lamproite magmas and melts related to them. Relative to their nature, until now there is no unified opinion [9, 14]. From the point of view of the complex of questions under discussion, these formations, probably, are close to the alkaline magmas, which transit stage of evolution passed to high-dynamic regime. In particular, the abundance of xenolith material in kimberlites indicates in our opinion that these fluid-silicate mixtures rose in the very high dynamic regime under mechanism of magma rupture or even fluid rupture. This regime caused sharp reduction in the strength of the surrounding rocks with their breakdown and capture by the ascending magma. Thus, different combinations of natural factors ensure the significant variety of the preeruptive magma evolution of different series. Multi-variability of the specific ways of 23
evolution causes the polygenesis of the magmas of these series. It is possible to assume that further solution of the problems touched upon in the work will make it possible to raise a question about the magmatism as non-linear phenomenon, like this was made by A.Shcheglov and I.Govorov in connection with polygenic ore deposits [23].
CONCLUSIONS The generalisation of theoretical developments and experimental research, which highlight different aspects of magmatic process in connection with geodynamics of CDZ, shows that the contemporary petrological and geodynamic models need the specific correction for the purpose of the proper evaluation of the whole number of possible modifications of the magma composition during their penetration into the upper levels of the Earth's crust, accessible to our direct study. Three main stages of the pre-eruptive magma evolution are distinguished: focus, transit and chamber. Each one is characterised by the independent set of the factors, which control directivity and scales of the modification of deep melts. The key factors of the evolution of magmas at the focus stage include: • • •
the isotherms uplift in combination with the decompression of some sections of the lithosphere caused by the mantle diapirs variations in composition and fluid content of mantle source degree of mantle heterogeneity.
In different CDZ these factors can make a different contribution to magmogenesis, causing the specific character of magmas already at the focus stage. Magma evolution at the transit stage is controlled by the following important factors: •
• • •
internal dynamic potential of the zones of magma genereation, that is: o excess pressure due to the volume effect of the reactions of melting o pressure, created by a difference in the densities between the melt and the surrounding rocks o pressure of the fluid dissolved in the melt rheological properties of melts regime of the volatile components of the ascending magma mechanical properties of the rocks of the lithosphere in CDZ.
Two main regimes of the magma uplift are distinguished: • •
with the insignificant crystallisation and the retention of contact with the surrounding rocks with the extensive crystallisation in the boundary parts and at the walls of channels.
Furthermore, the fluid saturation of the magmas can experience intensive crystallisation with melt boiling in the process of uplift. The chamber stage of the magma evolution is characterised by the dominant controlling role of fluid dynamic factors. They are determined by the large number of variables, accounting: • • •
rheological properties of magma dynamic parameters of magma motion geometric features of reservoirs
By this is ensured the multi-variability of the melt evolution in the periodically replenished transitional chambers, which is achieved within the framework of two different petrological processes: 24
• •
melt fractionation melt mixing.
Fractionation within the cells of the double-diffusive convection is periodically replacing by magma mixing as the properties of the lower and upper melts become closer. The lowspeed injections of the fresh portions of melts lead to the formation of the clearly stratified chambers with the autonomous fractionation in the different-level layers. High-speed injections can cause a certain mixing of the primary and differentiated magmas. Extensive hybridism can be achieved during magma withdrawal from the stratified chamber that compensates injections from below. The basalt-rhyolite chambers of this type and largescale hybridism of felsic and mafic melts having high value for CDZ, where the formation of the significant volumes of the rhyolitic melts of upper layer is possible as a result of the crustal anatexis. Transitional chambers are the unique filtering systems, which ensure preferred entering to the surface of the lightest melts and limit the possibility of penetration to the surface of dense primary magmas. However, the filtering role of chambers to a considerable extent depends on the dynamic regime of the injections of primary magmas from below. At some high-speed injections it is possible trans-chamber magma uplift without any melt conservation inside the differentiation reservoir. In conclusion, performed analysis has showed that practically at all stages of the preeruptive magma evolution in CDZ there are quite real possibilities of the differently directed modification of the composition of deep melts. Equal probability of participation in magmogenesis of different processes like partial melting, fractional crystallisation, hybridisma and crustal assimilation, capable of being occurred at different stages of preeruptive magma evolution, makes it possible to speak about the polygenesis of mantle magmas in CDZ. From this point of view the construction of the models of CDZ magmogenesis is possible on the basis of the sequential reconstruction of the modifications of the composition of the primary magmas, which occurred for each of the stages of preeruptive magma evolution.
REFERENCES 1. Belousov, A.F., Krivenko, A.P., 1983. Magmogenesis of Volcanic Formations. Novosibirsk, Nauka, 192 p. (In Russian) 2. Dobretsov, N.I., 1980. Introduction into Global Petrology. Novosibirsk, Nauka, 200 p. (In Russian) 3. Dowson, J., 1983. Kimberlites and their Xenoliths. Moscow, Mir, 300 p. (In Russian) 4. Lobach-Zuchenko, S.B., Krestin, E.M., Rybakov, S.I., et al., 1988. Greenstone belts of the Basement of Eastern European Platform. Leningrad, Nauka, 215 p. (In Russian) 5. Kadik, A.A., Frenkel, M.Y., 1982. Crustal and Upper Mantle Rocks Decompression as Magma Formation Mechanism. Moscow, Nauka, 120 p. (In Russian) 6. Lyashkevich, Z.M., 1987. Magmatism of Pripyat-Dneprovsk-Donetsk Paleorift. Kiev, Naukova Dumka, 176 s. (In Russian) 7. Bogatikov, O.A., Bogdanova, S.V., Borsuk, A.M., et al., 1987. Margmatic Rocks. V. 6. The Evolution of Magmatism in the History of the Earth. M., Science, 438 p. (In Russian) 8. Malyuk, B.I., 1987. On the dynamics of differentiation of basaltic magmas in the periodically refilled magma chambers (on the example of tholeiites of Precambrian greenstone belts). Geokhimia (Geochemistry International), N 8, pp.1084-1092. (In Russian) 9. Marakushev, A.A., 1982. Volcanic nature of the diamond-bearing rocks of kimberlitic pipes. Geology and Geophysics, N 8, p.16-27. (In Russian) 10. Milanovskiy, E.E., 1976. Rift Zones of Continents. Moscow, Nedra, 280 p. (In Russian) 11. Milanovskiy, E.E., 1983. Riftogenesis in the History of the Earth (riftogenesis in the anciane platforms). Moscow, Nedra, 280 p. (In Russian) 25
12. Ramberg, I., Morgan, P., 1984. Physical characteristic and the direction of evolution of continental rifts. In: 27 International Geological Congress. USSR, Moscow, 4-14 aug. 1984. Report volume 7. Tectonics. Moscow, Nauka, p.78-109. (In Russian) 13. Rast, H., 1972. Magma origin, uplift and emplacement. In: Mechanism of Magma Intrusion. Moscow, Mir, p.284-310. (In Russian) 14. Ryabchikov, I.D., 1980. Nature of kimberlitic "magmas". Geology of Ore Deposits, N 6, p. 18-26. (In Russian) 15. Salnikov, D.I., Traskin, V.Y., 1987. The basic ideas of physical chemistry geomechanics. In: Studies of Tectonic Deformations. Moscow, IFZ AN SSSR, p.33-83. (In Russian) 16. Seitov, N.S., 1987. Some aspects of the problem of rift genesis. Proceedings of Kazakhstan SSR AS. Ser. Geol., N 5, p. 35-44. (In Russian) 17. Sobolev, S.V., 1988. How the large volumes of magma can penetrate through the lithosphere. In: Composition and the Processes of the Deep Zones of Continental Lithosphere. Abstr.Internat.Symp. 30 May-2 June 1988, Novosibirsk, p.83. (In Russian) 18. Khlestov, V.V., 1980. The dynamics of the foci of melting in continuous solid medium and the evolution of mantle magmas. In: Mantle xenoliths and the problem of ultrabasic magmas. Abstr.All-Union Symp., 27-29 October, 1980, Novosibirsk, p. 11-12. (In Russian) 19. Khlestov, V.V., 1984. Dynamics of mantle magma formation in continental and oceanic segments of lithosphere. In: 27 International Geological Congress. USSR, Moscow, 4-14 aug. 1984. Abstracts. V. 9, Pt. 1, Moscow, Nauka, p.103-105. (In Russian) 20. Sharapov, V.N., 1976. To the dynamics of magmas intrusion. In: Replacement and Intrusion During the Magmatism and the Ore Formation. Novosibirsk, Nauka, p.103122. (In Russian) 21. Sharapov, V.N., Cherepanov, A.N., 1986. The Dynamics of the Magma Differentiation. Novosibirsk, Nauka, 90 p. (In Russian) 22. Shkodzinskiy, V.S., 1985. Magma Phase Evolution and Petrogenesis. Moscow, Nauka, 232 p. (In Russian) 23. Shcheglov, A.D., Govorov, I.N., 1985. Non-Linear Metallogeny and the Depth of the Earth. Moscow, Nauka, 324 p. (In Russian) 24. Allegre, C.J., Turcotte, D.L., 1986. Implications of a two-component marble-cake mantle. Nature, 323, p. 123-127. 25. Anderson, D.L., 1982. Chemical composition and evolution of the mantle. Adv. Earth Planet. Sci., 12, p.301-318. 26. Baker, B.H., McBirney, A.R., 1985. Liquid fractionation. Part III: Geochemistry of zoned magmas and the compositional effects of liquid fractionation. Processes in Magma Chambers. J. Volcanol. Geotherm. Res., 24, p.55-81. 27. Bickle, M.J., Ford, C.E., Nisbet, E.G., 1977. The petrogenesis of peridotitic komatiites: evidence from high-pressure melting experiments. Earth Planet. Sci. Lett., 37, p.97-106. 28. Blake, S., 1981. Eruption from zoned magma chamber. J. Geol. Soc. London, 138, p. 281-287. 29. Blake, S., 1987. Sediment entrainment in viscous fluids: can crystals be erupted from magma chamber floors ? J. Geol., 95, p.397-406. 30. Blake, S., Campbell, I.H., 1986. The dynamics of magma-mixing during flow in volcanic conduits. Contribs. Mineral. Petrol., 94, p.72-81. 31. Blake, S., Ivey, G.N., 1986. Magma-mixing and the dynamics of withdrawal from stratified reservoirs. J. Volcanol. Geotherm. Res, 27, p. 153-178. 32. Campbell, I.H., 1978. Some problems with cumulus theory. Lithos, 11, p.311-323. 33. Campbell, I.H., 1985. The difference between oceanic and continental tholeiites: a fluid dynamic explanation. Contribs. Mineral. Petrol., 91, p.37-43. 34. Campbell, I.H., Turner, J.S., 1985. Turbulent mixing between fluids with different viscosities. Nature, 313, p.39-42. 35. Campbell, I.H., Turner, J.S., 1986. The influence of viscosity on fountains in magma chambers. J. Petrol, 27, p.1-30. 26
36. Chen, C.P., Turner, J.S., 1980. Crystallisation in a double-diffusive convective system. J.Geophys.Res, 1980, 85, p.2573-2593. 37. Cox, K.G., 1978. Komatiites and other high-magnesia lavas: some problems. Phil. Trans. Roy. Soc. London, A288, p.599-609. 38. Crisp, J.A., 1984. Rates of magma emplacement and volcanic output. J. Volcanol. Geotherm. Res., 20, p. 177-211. 39. Daly, R.A., 1903. The mechanics of igneous intrusion. Amer.J.Sci., 16, p.107-126. 40. Delaney, P.T., Pollard, D.D., 1982. Solidification of basaltic magma during flow in a dike. Amer.J.Sci., 282, p.836-885. 41. Emerman S.H., Turcotte, D.C., Spence, D.A., 1986. Transport of magma and hydrothermal solutions by laminar and turbulent fluid fracture. Phys. Earth. Planet. Interior, 41, p.249-259. 42. Huppert, H.E., 1986. The intrusion of fluid mechanics into geology. J. Fluid Mech., 173, p.557-594. 43. Huppert, H.E., Sparks, R.S.J., 1980. The fluid dynamics of a basaltic magma chamber replenished by influx of hot, dense ultrabasic magma. Contrib. Mineral. Petrol, 75, p.279-289. 44. Huppert, H.E., Sparks, R.S.J., 1984. Double-diffusive convection due to crystallization in magmas. Ann. Rev. Earth Planet. Sci., 12, p.11-37. 45. Huppert, H.E., Sparks, R.S.J., Whitehead, J.A., Hallworth, M.A., 1986. Replenishment of magma chambers by light inputs. J. Geophys. Res., 91, p.61136122. 46. Huppert, H.E., Sparks, R.S.J., Wilson, J.R., Hallworth, M.A., 1986. Cooling and crystallization at an inclined plane. Earth Planet. Sci. Lett., 79, p.319-328. 47. Huppert, H.E., Turner, J.S., 1981. Double-diffusive convection. J. Fluid. Mech., 106, p.299-329. 48. Huppert, H.E., Turner, J.S., Sparks, R.S.J., 1982. Replenished magma chambers: Effects of compositional zonation and input rates. Earth Planet. Sci. Lett., 57, p.345-357. 49. Keen, C.E. 1985. The dynamics of rifting: deformation of the lithosphere by active and passive driving forces. Geophys. J. Roy. Astron. Soc., 80, p.95-120. 50. Kirby, S.H., 1980. Tectonic stresses in the lithosphere: constraints provided by experimental deformation of rocks. J. Geophys. Res., 85, p.6353-6363. 51. Huppert, H.E., Sparks, R.S.J., Wilson, J.R., Hallworth, M.A., Leitch, A.M., 1987. Laboratory experiments with aqueous solutions modelling magma chamber processes. II. Cooling and crystallization along inclined planes. In: I.Parsons (Ed) Origin of Igneous Layering. Reidel PC, p.539-568. 52. Maaloe, S., Scheie, A., 1982. The permeability-controlled accumulation of primary magma in a planetary mantle. Phys. Earth Planet. Inter., 29, p.344-353. 53. Marsh, B.D., 1982. On the mechanics of igneous diapirism, stoping and zone melting. Amer. J. Sci., 282, p.808-855. 54. McBirney, A.R., Baker, B.H., Nilson, R.H., 1985. Liquid fractionation. Part I: Basic principles and experimental simulations. In: Processes in Magma Chambers. J. Volcanol. Geotherm. Res., 24, p.1-24. 55. McBirney A.R., Noyes, R.M., 1979. Crystallisation and layering of the Skaergaard intrusion. J. Petrol., 20, p.467-554. 56. McKenzie, D., 1984. The generation and compaction of partially molten rocks. J. Petrol., 25, p.713-765. 57. Nilson, R.H., McBirney, A.R., Baker, B.H., 1985. Liquid fractionation. Part II: Fluid dynamics and quantitative implications for magmatic systems. In: Processes in Magma Chambers. J. Volcanol. Geotherm. Res., 24, p.25-54. 58. O'Hara, M.J., 1985. Importance of the "shape" of the melting regime during partial melting of the mantle. Nature, 314, p. 58-62. 59. Richter, F.M., McKenzie, D., 1984. Dynamical models for melt segregation from a deformable matrix. J. Geol., 92, p.729-740. 60. Scott, D.R., Stevenson, D.J., 1986. Magma ascent by porous flow. J. Geophys. Res., 91, p.9283-9296.
27
61. Shaw, H.R., 1980. The fracture mechanism of magma transport from the mantle to the surface. In: R.B.Hargraves (Ed.) Physics of Magmatic Processes. Princeton, N.J., Princeton Univ. Press., p.201-264. 62. Spence, D.A., Turcotte, D.L., 1985. Magma driven propagation of cracks. J.Geophys. Res., 90, p.575-580. 63. Spera, F.J., 1980. Aspects of magma transport. In: R.B.Hargraves (Ed.) Physics of Magmatic Processes. Princeton, N.J., Princeton Univ. Press., p.265-314. 64. Spera, F.J., 1984. Some numerical experiments on the withdrawal of magma from crustal reservoirs. J. Geophys. Res., 89, p.8222-8236. 65. Spera, F.J., 1987. Dynamics of translithospheric migration of metasomatic fluid and alkaline magma. In: Mantle Metasomatism. - London: Academic Press, p.1-20. 66. Spera, F.J., Yuen, D.A., Greer, J.C., Sewell, G., 1986. Dynamics of magma withdrawal from stratified magma chambers. Geology, 14, p.723-726. 67. Stolper, E., Walker, D., 1980. Melt density and the average composition of basalts. Contrib. Mineral. Petrol., 74, p. 7-12. 68. Takahashi, E., 1986. Melting of a dry peridotite KLB-1 up to 14 Gpa: implications on the origin of peridotitic upper mantle. J. Geophys. Res., 91, p.9367-9382. 69. Turcotte, D.L., Ahern, J.L., 1978. A porous flow model for magma migration in the astenosphere. J. Geophys. Res., 83, p.767-772. 70. Turner, J.S., Campbell, I.H., 1986. Convection and mixing in magma chambers. // Earth-Sci. Revs, 23, p.255-352. 71. Turner, J.S., Gustafson, L.B., 1981. Fluid motions and compositional gradients produced by crystallization or melting at vertical boundaries. J. Volcanol. Geotherm. Res., 11, p.93-125. 72. Weertman, J, 1971. Theory of fluid-filled crevasses in glaciers applied to vertical magma transport beneath ridges. J. Geophys. Res, 75, p. 1171-1183. 73. Wickham, S.M., 1987. The segregation and emplacement of granitic magma. J. Geol. Soc. London, 144, p.281-297.
© 1989, Malyuk, B.I. Factors of Pre-Eruptive Magma Evolution in the Continental Destructive Zones Institute of Geology & Geochemistry of Combustible Minerals Preprint # 89-1, 56 p. (In Russian) © 2001 - Web adaptation by B.I.Malyuk at former website of the Institute of Geology and Geochemistry of Combustible Minerals, Ukrainian National Academy of Sciences, Lviv, Ukraine English translation via BabelFish
Original front page shortcut
28
.
