fluids in the Earth

fluids in the Earth One dictionary definition of a fluid is ‘a substance that flows freely’. This definition is, however, more general than is commonly used in the Earth sciences, for it would include silicate melts, or even the solid mantle, since it too flows, albeit extremely slowly. Some authors use the term ‘fluids’ for common gases and liquids (such as water, water vapour, carbon dioxide, etc.); others restrict the noun ‘fluid’ to volatile substances at temperatures above their critical temperatures. Regardless of the terminology, fluids are understood to be distinct from the denser silicate melts (and of course the solid silicates) and generally to have compositions in common with substances that are gases or liquids at the Earth's surface. Obvious examples are water, methane, carbon dioxide, carbon monoxide, and sulphur dioxide.

Although there are geological systems in which free fluid phases exist or have existed in the past, there are also many processes and systems in which fluids are important but do not form their own separate phase. Rather they are incorporated into minerals and melts, either making distinct minerals or present as defects. While it is then not strictly correct to describe them as fluids (since they are not capable of flowing freely), this is commonly done. References are often made to the water content of a magma, even though there is no water in the sense of a free water phase, nor may there be any H₂O molecules in the magma. This common usage comes about because when heated to above saturation, oxygen and hydrogen in the melt or mineral which exist as OH groups can dehydrate to form bubbles of water. It is more strictly correct to refer to these components as volatiles rather than fluids, but the terms are commonly used synonymously.

The chemical composition of fluids in the Earth

The predominant chemical components of most fluids in the Earth are carbon, hydrogen, oxygen, and sulphur. According to the conditions (temperature, pressure, and oxygen partial pressure) and composition, these elements will form molecular species such as CO, CO₂, CH₄, H₂, H₂S, SO₂, etc. In addition they are able to dissolve large concentrations of many elements, particularly at higher temperatures. At the low temperatures and pressures of the Earth's surface, water will readily dissolve large amount of salts, such as sodium chloride (NaCl), but the solubility of metals is extremely low. At high temperatures (500 °C or more), however, metal concentrations in water–NaCl fluids can reach 1 per cent by weight (1 wt%) or more. Clearly such fluids have been important in ore formation.

How much fluid is there, and where is it?

Taking the broader definition of a fluid as being the volatile component of the systems rather than a free fluid phase, estimates for the fluid content of the whole Earth are very variable and are susceptible to large uncertainties. An analysis of meteorites suggests that the water content of the whole Earth could be as high as 2 or 3 per cent by weight and the carbon dioxide (CO₂) content even higher. These amounts are far greater than estimates for the current volatile content of the Earth. The reason for the discrepancy is twofold. First, if, as is suggested, the Earth was hit by a Mars-sized object early in its history, then a large proportion of the volatile fluids would have been lost to space during this collision. Secondly, the Earth has been continually degassing since its formation, although much of this happened early and the rate is now far slower. This means that we cannot rely on meteorites; instead we have to calculate the total Earth fluid budget by summing up the different reservoirs: the hydrosphere, crust, mantle, and core. It should be recognized that estimates made this way can differ significantly, primarily because of uncertainties in the estimates for the mantle and, in the case of carbon and sulphur, the core.

Estimates for the hydrogen content of the whole Earth are around 30 ppm by weight, which amounts to an equivalent of about 300 ppm (parts per million) water. This is much more water than is in the oceans alone, with much of it contained in the Earth's crust and mantle, and perhaps also the core. In environments very near the surface, water concentrations in aquifers can reach 50 to 60 per cent by volume, but this rapidly falls off with depth since the pressure forces the fluid out of the pores. At greater depths and higher temperatures, the water reacts with the host rock to create hydrous phases such as clays, serpentines, amphiboles, and micas. In the Earth's mantle the water content is thought to be in the region of 100 ppm. This is estimated from the water content of mid-ocean-ridge basalts (MORBs), which typically contain about 3000 ppm water, and assuming that MORBs come from about 10 per cent melting of the upper mantle. Some rare MORBs, however, contain significantly more water. The water in the mantle may be contained either in high pressure hydrous phases, or alternatively as small hydrous defects in norminally anhydrous minerals. Olivine, the most abundant mineral in the upper mantle, can contain 1000 ppm water, and perhaps significantly more. Wadsleyite, a dense magnesium silicate thought to exist in the Earth's transition zone (400 to 550 km depth), is capable of containing up to 3 wt% water. This has led to suggestions that this part of the Earth's interior might be a large reservoir for water.

Estimating the amount of carbon in the whole Earth is also problematic, and estimates range from 30 to 450 ppm by weight. When converted to CO₂ this is about 100 to 1500 ppm. One of the problems in estimating the carbon content of the Earth is that there may be considerable amounts in the core. If we consider mole per cent instead of weight per cent, the concentrations of carbon and hydrogen in the Earth are similar. The bulk of CO₂ in the near surface is contained in carbonate rocks, primarily limestones. Estimates for mantle CO₂ abundances are more difficult to obtain than for water, since MORBs almost always show evidence of CO₂ loss. Nevertheless, estimates of mantle CO₂ are about 300 ppm. Mantle CO₂ may be primarily contained in the high-pressure magnesium carbonate phase, magnesite (MgCO₃), although it is very rare to find carbonates in mantle xenoliths. Another likely possibility is that the majority of the mantle carbon is in the form of diamond (C).

As with carbon, sulphur is also an element that could be in the core. Estimates for the amount in the core range from 0 to 9 wt% and have a strong influence on estimates for whole Earth concentrations. In the silicate part of the Earth (the crust and mantle), sulphur concentrations are about 250 ppm, with similar amounts in the crust and mantle. This is about 1000 ppm (sulphur dioxide (SO₂). Sulphur in the mantle is assumed to be in the form of iron–nickel sulphides. These are fairly widespread in mantle xenoliths.

The effect on the physical properties of minerals and melts

The presence of small amounts of fluid species such as H₂O and, to a lesser extent, CO₂ can have significant effects on many physical properties of minerals and melts. Concentrations of water as low as 100 ppm can increase creep rates in minerals and melt by more than two orders of magnitude. Hydrous phases can be greater than 10 per cent more compressible than their anhydrous equivalents. Seismic velocities are lowered and attenuation is also increased by the presence of water. Similarly, electrical conductivity can be strongly increased or decreased by small amounts of water. Perhaps one of the most important effects is that melting temperatures of rocks and minerals are strongly affected by water, with liquidus temperatures being lowered by many hundreds of degrees. This is an extremely im-portant effect and has implications in many different environments.

Fluids in metamorphism

The majority of metamorphic reactions typically involve fluids in some way or another. This can be either by adding fluids, as in hydrothermal alteration, or by releasing fluids during decarbonation or dehydration reactions. Apart from changing the equilibrium phases in the rock, fluids can also change the chemical composition by adding and taking away chemical components.

Fluids in the lower crust

Magnetotellurie studies of the electrical conductivity of the lower crust show that it is anomalously conducting. This has led to the suggestion that the lower crust is saturated with a film of brine. There is, however, an alternative view which argues that the lower crust cannot be wet and that we must look for another mechanism for the high conductivity. This alternative view argues that at the temperatures and pressures of the lower crust, any free fluid would react with the rock to form hydrous mineral phases.

Fluids in volcanism

Fluids dissolved in magmas are the main cause of explosive volcanism. If the fluid reaches saturation in the magma, it then begins to exsolve. If this happens slowly, the gas bubbles will slowly migrate through the magma and degas. If, on the other hand, the gas exsolves quickly, the rapid and very large increase in the volume of the fluid as it enters the gas phase can generate tremendous mechanical energy and cause the eruption to be explosive. This can come about if the pressure in the magma chamber is suddenly released or if the magma has a high viscosity that inhibits migration of the bubbles. Water release from volcanic eruptions through geologic time has produced the Earth's oceans and atmosphere.

Fluids and earthquakes

For an earthquake to occur, the stress has to exceed the strength of the rock. One of the ways in which this can be achieved is if the fault is lubricated by a fluid. This has the effect of decreasing the friction. Dehydration reactions release water and increase the total volume of the rock, thereby pushing apart the grains and removing some of the strong frictional forces.

Dehydration reactions may also be necessary for deep subduction-zone earthquakes to occur. Many earthquakes in subduction zones occur at depths at which the pressures are so large that the friction between two sides of a fault will stop the fault from sliding; the stress should then be released by creep rather than by faulting. It has been suggested that release of water during the dehydration of serpentine can effectively lubricate a fault.

Water in subduction zones

It is generally accepted that water plays a crucial role in the generation of melting and volcanism beneath island arcs, where cold oceanic lithosphere is subducted into the mantle. The altered basaltic layer in the ocean crust typically contains about 3 per cent H₂O; as the lithosphere subducts and heats up, the hydrous phases in the slab dehydrate and eventually release their fluids. These fluids migrate into the overlaying dry mantle, which causes the liquidus temperature of these rocks to be lowered. At some point some of the mantle melts and migrates to the surface, where it forms island-arc volcanoes. There is some evidence that the amount of melting is correlated with the amount of water coming from the slab. Without the water, the cold subducting slab would simply cool the surrounding mantle and there would be no arc volcanism.

Recycling of fluids into the mantle

Mid-ocean ridge basalts show clear evidence for small amounts of fluids being degassed from the mantle. Although subduction zones return fluids to the mantle, much of that fluid then returns to the surface in island-arc volcanism. The question whether any fluid continues past the melting zone into the deeper mantle remains contentious. It has been suggested that in subduction zones where old, cold lithosphere is being subducted, the mantle near the slab would be too cold to melt and water might either react with it to form hydrous phases or be incorporated as defects in mantle olivines or pyroxenes. It might then be carried down to the transition zone, where it could readily be incorporated into wadsleyite, and then deeper, into the lower mantle.

John P. Brodholt