Where is Mars’s water?

A delta at the edge of Jazero Crater on Mars; definite evidence that water once flowed into the crater. Colours show different minerals in the delta sediments (credit: Brown University)

Early in the exploration of Mars using orbiting imaging systems it was easy to be sceptical about evidence for water being present at or near the surface of the Red Planet. Resolution was poor and some claims seemed to be wishful thinking or a sort of astronautical agitprop. For instance, gullies on steep slopes appeared so sharp that they must be forming continually, otherwise Mars’s periodic huge dust storms would have muted them. Some scientists claimed that they were signs of flowing water and even presented pictures from different overpasses that showed changes in them, such as darkening and small shifts in microtopography, which may have resulted from flowing water. Because Mars has a mean surface temperature of about -50°C that seems unlikely; at such extremes in Antarctica spit at the ground and it lands as ice. Nonetheless a bit of special pleading that deeply buried ice in Martian sediments might melt because of pressure gave the idea some traction.

A far more plausible explanation for the active gulley formation is that loose fine sediment can flow in the manner of a liquid, as it does in sand dunes on Earth (see: First signs of liquid water on Mars? June 2000). Yet as remotely sensed image coverage expanded and its resolution improved (currently about 50 cm) masses of evidence for drainage networks, signs of catastrophic floods and even glaciers (The glaciers of Mars, July 2003) emerged. Huge areas of the planet bore witness to a period in its past history – 4.1 to 3.8 billion years (Ga) ago – when it was a warm and wet planet. It has even been suggested that the flat, low-elevation northern hemisphere was the bed of a former ocean, covering about a third of Mars to a depth of about a kilometre. Now the planet has a hyperarid surface and a very thin atmosphere dominated by CO2, a little nitrogen and argon but almost no water vapour (~0.03%). Its poles are covered by ice caps whose extents fluctuate seasonally. They each have a core of permanent water ice, and seasonally expand and contract due to formation and sublimation of dry ice made of solid CO2. So what happened to Mars’s once abundant water?

One long-held theory is that water and most of Mars’s original atmosphere escaped to space. A suggested mechanism is the photo-dissociation of water to hydrogen and oxygen. Mars’s gravity cannot prevent hydrogen escape, which would leave an excess of atmospheric oxygen. One thing in abundance on the Martian surface is oxygen combined in iron oxides (Fe2O3); hence its red coloration. This hematite may have formed during chemical weathering of surface rocks and sediments during the wet phase, which released Fe2+ ions that were immediately oxidised by the hyper-oxygenated atmosphere that resulted from photo-dissociation. But there is another plausible explanation …

The lake-bed sediments of Gale Crater on Mars from NASA’s Curiosity rover (credit: NASA/JPL, California Institute of Technology)

The much publicised successful landing of NASA’s Perseverance rover on 18 February 2021 was aimed at the small Jezero Crater, near the Martian equator. This contains an indisputable delta of a large drainage system that must once have filled the crater with a circular lake; a good place to seek out signs of early life, for which Perseverance is impressively equipped. Shortly afterwards there appeared a Research Article in Science (Scheller, E.L. et al. 2021. Long-term drying of Mars by sequestration of ocean-scale volumes of water in the crust. Science, Online research article eabc7717; DOI: 10.1126/science.abc7717) that examines the fate of the planet’s water. The authors estimate that by 3.0 Ga Mars’s surface had reached its current dry state. They model three processes – supply of water by volcanic degassing and its loss by atmospheric escape and chemical weathering of the Martian surface. The modelling was constrained by the ratio of deuterium (2H) to hydrogen inferred from meteorites believed to come from Mars and estimates by orbiting spacecraft of the current escape of hydrogen from the atmosphere. The latter is too slow to explain the huge loss of water between 4 and 3 Ga and subsequently. Addition of water from Mars’s mantle by volcanoes, even from the gigantic Olympus Mons, was far slower than on Earth because continuous plate tectonics was never achieved on Mars. Chemical weathering of the surface during Mars’s warm-wet phase formed abundant hydrated minerals as well as the hematite that gives the planet its characteristic hue. Water transport before 3 Ga moved clays and hydroxides etc to sedimentary basins, where they have remained undisturbed. On Earth, tectonics recycles sediments and their content of hydrated minerals into the mantle, eventually to regurgitate their water content through volcanism. On Mars, weathering and deposition has irreversibly locked-up between 30 and 99% of Mars’s original endowment of water in its ancient sedimentary crust.

That seems to be a ‘bit of a downer’ for ambitious prospects of terraforming Mars and making it a human escape destination. There are, however, some locations where water may be available in sufficient quantities to support some kind of permanent presence of small colonies, in the form of buried layers of ice, similar to permafrost (see: Ice cliffs on Mars, January 2018)

See also:  Carr, M.H. 2012. The fluvial history of Mars. Philosophical Transaction of the Royal Society (A), v. 370, p. 2193-2215; DOI: 10.1098/rsta.2011.0500.

Is there water in the Earth’s core?

Understandably, the nature of what lies at the centre of the Earth is as much the subject of speculation as tangible evidence. That there must be something very dense within the planet emerged once the Earth’s bulk density was calculated. Because a high proportion of meteorites are dominated by an alloy of the metals iron and nickel, geoscientists adopted that combination as plausible core material. Study of the arrival times around the globe of seismic waves from earthquakes then revealed the actual size of the Earth’s core. Iron-nickel alloy fitted the bill quite nicely. It also fits geochemical evidence, such as the crust and mantle’s depletion in some trace elements that theoretically have an affinity for iron. The fact that seismology showed also that the outer core was molten and able to flow, together with metals’ high electrical conductivity, gave rise to the current concept of the geomagnetic field being generated by a dynamo effect in the core. However the density of Fe-Ni is not ‘quite right’ because the core is somewhat lighter than predicted for the pure alloy under stupendous pressure: it must contain a substantial amount – up to 13% – of lower density materials.  Silicon, sulfur and oxygen have been suggested as candidates, with evidence from a variety of minor minerals in metallic meteorites.

A recent model for core formation (credit Crystal Y. Shi et al 2013; DOI: 10.1038/NGEO1956 Fig. 5)

The world is currently awash with models that attempt to throw light on the course of the Covid-19 pandemic. Many are based on highly uncertain data, leading to suggestions by some people that they have become tools for political elites and a means of helping ambitious scientists into the limelight: a sort of fuel for hubris. In the midst of this unprecedented turmoil there has appeared a suggestion (from modelling) that the core also contains abundant hydrogen (Li, Y. et al. 2020. The Earth’s core as a reservoir of water. Nature Geoscience, v. 13, published online; DOI: 10.1038/s41561-020-0578-1). Yunguo Li and colleagues, from University College London, the Chinese Academy of Science and the University of Oslo, explore the idea that the dominant hydrogen of the pre-planetary Solar nebula, which accreted to form the Earth, may have joined iron during core formation. This had been predicted from the thermodynamics of chemical reactions between water and iron. The team takes this further through the geochemical theory that elements and compounds tend to enter other materials preferentially. For example, during partial melting of the crust alkali metals (Na, K etc) are more likely to enter the granitic melt than to remain in the solid residue. Li et al. have used thermodynamics to predict the partitioning of hydrogen between iron and silicate melts under the very high temperature and pressure conditions at the boundary between the core and mantle.

Their calculations suggest that hydrogen then behaves in much the same manner as, say gold and platinum: it becomes ‘iron-loving’ or siderophile and prefers the molten core, as would H2O. The amount that gets in depends on the water content of the molten silicate that eventually becomes the mantle. If the water now making up Earth’s ocean was ‘degassed’ from the mantle during core formation then the original silicate melt would have been ‘wetter’ than it is now. The implication of such early degassing is that the core may contain 5 ‘oceans worth’ of water! The alternative scenario for Earth’s becoming a watery world is the later accretion of, for instance, cometary material. In that case, the early core would have been drier. Yet, the continual subduction of hydrated oceanic lithosphere into the deep mantle during billions of years of plate tectonics would steadily have added water to the core, in the form of iron oxides and hydrogen. So, the core might, in either case, contain several ‘oceans’ of the components of water. One line of indirect evidence is the deficiency in Earth’s actual water of the heavier isotope of hydrogen (deuterium) relative to the D/H ratio of chondritic meteorites. Theory suggests that D has slightly more affinity for joining iron than does H. Substantial water in the core does help explain the core’s apparent low density, but that notion requires as much faith as politicians seem to have in ‘following the Science’ during the current pandemic …