Experimental petrology of basalts and their source rocks

Abstract

This volume is concerned with the basaltic effluent from planetary interiors, and this chapter is concerned with the extent to which experimental petrology of basalts can inform us about the composition and mineralogy of the source material within the interiors of planets. If the compositions of sources can be deduced using the experimental petrology of basalts, or in any other way (Chapter 4), then the relationship between source rock and basaltic magma can be investigated by complementary experiments using the proposed source material. Basaltic volcanism appears to be a characteristic feature of most investigated planetary surfaces (Chapters 2 and 5), with basalts being produced either when internal temperatures become high enough to cause partial melting of the planetary rock, or when impacts by other bodies raise the temperature high enough to cause partial or complete melting of near-surface rocks. The main variables to be considered in these processes are the pressure and equivalent depth, the temperature, and the composition and mineralogy of the planetary interior and near-surface rocks. Of these three variables, the value of pressure as a function of depth is the best known. A given pressure is achieved at depths that vary considerably from one planetary body to another as illustrated in Fig. 3.1.1. Temperature is a variable that changes as a function of time, and as a function of process (Chapter 9). Convective movements within a planet, for example, influence both the rate of heat transport to the surface, 0 and the development of regional variations in temperature distribution versus depth. The temperature produced by meteoritic or planetesimal bombardment depends upon the mass, velocity, and frequency of impacting bodies. Variations reviewed in Chapter 9 indicate the extent of our uncertainty about the temperatures of planetary interiors, even for our own planet. Figure 3.1.2 shows three geotherms that have been proposed for a convecting mantle within the present Earth (Solomon, 1976). The geotherms drawn for the lithosphere (shallower than about 100 km) are passed through the zones plotted by Solomon ( 1976) for values estimated from study of peridotite nodules from the mantle; these geotherms are similar to those calculated for conduction models (e.g., Clark and Ringwood, 1964). Temperatures in a region of upwelling, beneath ocean ridges for example, are higher than temperatures beneath normal ocean plates. The temperatures shown beneath shield and ocean plates become identical at 200 km depth (contrast Jordan, 1975). The third variable, the composition and mineralogy of basalt source regions, is what we seek to determine from experimental petrology of erupted basalts. For the Earth, however, we already have a fairly well-defined model for the structure and petrology of the mantle, based on cosmochemistry, geophysics, and petrology (Chapter 4). The composition corresponds to that of a peridotite dominated at low pressures by olivine and two pyroxene minerals. Figure 3.1.2 outlines the main phase relationships for this material up to pressures of 250 kb, corresponding to a depth of 700 km within the Earth. For other planetary bodies dominated by the components of olivine and pyroxenes, the phase relationships would be similar, but the specific depth scales would differ (Fig. 3.1.1 ). Basaltic magmas are generated within the melting interval (crystals + liquid), when the temperature becomes high enough to cross the solidus. According to the terrestrial geotherms given in Fig. 3.1.2, melting does not occur at all beneath shield and ocean plates, and therefore some special circumstance is required to increase the temperature locally to account for the oceanic volcanoes and the continental flood basalts. It is generally assumed that upward convection of mantle material raises the temperature at relatively shallow levels, as indicated by the geotherm associated with mantle upwelling beneath ocean ridges in Fig. 3.1.2. In this tectonic environment, a geotherm crosses the solidus at depths of 70 ± 40 km (sections 3.3.7 and 3.3.8). Each planetary body has characteristics that dis- , tinguish it from the others, even if it should turn out that most of them have similar major element compositions. The different pressure-depth relationships suggest that temperature distribution curves are likely to intersect solidus curves at different pressures on the different bodies. The phase relationships shown in Fig. 3.1.2 may be changed by variations in Fe, Mg, Ca, and Al proportions, and changed radically by the addition of volatile components (sections 3.3.2 and 3.3.3). For parts of the Earth's interior containing traces of CO_2 and H_2O, there is a wide interval of incipient melting below the solidus plotted in Fig. 3.1.2. For the Moon, in contrast, the concentration of volatile components appears to be negligible. The oxygen fugacity of lunar rocks is much lower than that of terrestrial rocks. Mars may be enriched in Fe and S compared with Earth and Moon (McGetchin and Smyth, 1978). The occurrence of partial melting and volcanism depends upon the maintenance of sufficiently high temperatures and thus upon the body's thermal history (Chapter 9). Parent bodies of basaltic achondrite meteorites are believed to have solidified at an early stage in the development of the solar system, and active volcanism on the Moon ceased relatively early in its history. Mars appears to have ceased evolving, and the Earth continues with active volcanism caused to a large extent by the mass movement associated with plate tectonics. Therefore, in considering the basaltic volcanism of each planetary body, we have different characteristics, and different ground rules for interpretation. The spectacular eruptions on Io were not anticipated according to preexisting ground rules. The purpose of this chapter is to outline the methods used to determine whether a particular basalt is likely to contain direct information about its source region by application of the methods of experimental petrology. In addition, the experimental work on terrestrial, lunar, and meteoritic basalts, as well as likely source materials, will be reviewed to show the extent to which the link between the basalts and their source regions has been established, and to show the extent to which additional processes are responsible for the chemistry observed in the basalts. We will conclude that experimental petrology does provide an internally consistent framework for understanding basalts as the melting products of ultrabasic mantle assemblages, and that laboratory experiments can be used for the exercise of going back from basalt chemistry to source region constitution, provided that certain conditions are satisfied.

Files

Additional details