Mercury is the closest planet to the Sun. At its equator in daytime, the surface can reach about 430 degrees Celsius, well above the melting point of lead.<\/p>\n
Near the poles, however, the floor of a deep crater can remain below minus 170 degrees Celsius. Water ice lies at the surface in the coldest of these places and beneath a dark insulating layer in others.<\/p>\n
I find the contrast easier to understand once the word \u201cnear\u201d is handled carefully. Mercury has ground hot enough to melt lead, but not every sunlit patch beside a polar crater reaches that temperature. The hottest figures refer mainly to lower latitudes under direct sunlight. At the poles, sunlight arrives at a shallow angle.<\/p>\n
Even so, the boundary between illuminated rock and permanent shadow can be abrupt. With almost no atmosphere to circulate heat, a crater rim and the floor below it can occupy very different thermal environments.<\/p>\n
NASA now says its\u00a0MESSENGER mission verified that Mercury\u2019s polar deposits are dominantly water ice<\/a>. That is a reasonable summary of a case built in 2012 from three independent kinds of evidence. It does not mean the spacecraft landed, drilled or held a piece of Mercurian ice in an instrument.<\/p>\n Permanent shadow is a problem of geometry<\/p>\n Mercury\u2019s rotation axis is tilted by only a small fraction of a degree. The Sun therefore stays close to the horizon at the poles rather than climbing high across the sky with the seasons. In a sufficiently deep high-latitude crater, the rim blocks direct sunlight throughout Mercury\u2019s year.<\/p>\n \u201cPermanent\u201d describes the illumination produced by the planet\u2019s present spin and the crater\u2019s topography. It should not be read as proof that no ray of sunlight has reached a particular floor at any time since Mercury formed. Craters have ages, and the planet\u2019s orientation has a history. What researchers can map is terrain that receives no direct sunlight under the relevant geometry today.<\/p>\n The lack of substantial air is just as important. Mercury has an exosphere, an extremely sparse population of atoms released from the surface, rather than an atmosphere capable of carrying heat around the planet.<\/p>\n That makes Mercury a useful contrast with Venus. In\u00a0my recent article on possible microbial survival in Venus\u2019s clouds<\/a>, the atmosphere was central to why temperature and pressure change with altitude. Mercury offers the opposite arrangement. The temperature of a piece of ground depends strongly on whether sunlight and thermal radiation can reach that particular piece of ground.<\/p>\n The shadow is not perfectly isolated. Crater walls can scatter light and radiate heat towards the floor, while the regolith conducts some heat from nearby material. Researchers therefore calculate temperatures rather than treating every dark pixel as equally cold.<\/p>\n Radar found the deposits before MESSENGER explained them<\/p>\n The first strong clue did not come from a spacecraft. In 1991, observations using the Arecibo radio telescope found unusually bright radar patches near Mercury\u2019s north pole. Similar features were later mapped in the south.<\/p>\n Water ice can produce this radar behaviour, but it was not the only proposed material. Sulphur and rough surface textures were among the alternatives. Mariner 10 had photographed less than half of Mercury in the 1970s, so scientists could not yet match every bright radar feature to a detailed map of polar craters.<\/p>\n MESSENGER changed that after entering orbit in March 2011. Its cameras mapped the polar topography and illumination over time. By 2012, the mission team found that the radar-bright deposits at both poles were confined to permanently shadowed terrain.<\/p>\n Location alone was not enough. The stronger case came from three studies published together in\u00a0Science.<\/p>\n First, MESSENGER\u2019s neutron spectrometer measured fewer energetic neutrons escaping from the north polar region. Hydrogen is particularly effective at slowing neutrons, and the size and distribution of the signal were best explained by water-rich material.\u00a0David Lawrence and colleagues estimated<\/a>\u00a0a total polar water mass between about 2.1 \u00d7 1013\u00a0and 1.4 \u00d7 1015\u00a0kilograms, a deliberately broad range.<\/p>\n Second, the spacecraft\u2019s laser altimeter found bright surfaces in some of the very cold areas and unusually dark surfaces in somewhat warmer ones.\u00a0Gregory Neumann and colleagues interpreted<\/a>\u00a0the bright deposits as exposed ice and the dark deposits as a covering above buried ice.<\/p>\n Third,\u00a0thermal models led by David Paige<\/a>\u00a0predicted where surface ice and buried ice should remain stable. Those predicted zones matched the radar and reflectance patterns.<\/p>\n No one of those measurements would have carried the same force alone. Together they connected a radar signature, a hydrogen signal, surface brightness and temperatures calculated from the terrain.<\/p>\n Some ice is exposed and some is under a dark cover<\/p>\n The coldest crater floors closest to the north pole have modelled maximum surface temperatures below 100 kelvin, or about minus 173 degrees Celsius. NASA\u2019s\u00a0temperature map of the north polar region<\/a>\u00a0shows exposed ice to be stable in places such as Prokofiev crater.<\/p>\n Farther from the pole, a crater floor may still avoid direct sunlight but receive enough indirect heat to make exposed water ice slowly escape into space. In these locations, MESSENGER found dark material where thermal models said ice could survive only underground.<\/p>\n The proposed covering is thin on a planetary scale, perhaps tens of centimetres, but enough to reduce loss from the ice below. It may contain complex organic compounds delivered with water by comets or volatile-rich asteroids.<\/p>\n \u201cOrganic\u201d here is a chemical category, not evidence of organisms. Carbon-bearing compounds can form without life and are found in meteorites and comets. Mercury\u2019s polar ice is also not a liquid habitat. It is a frozen deposit in a near-vacuum, exposed to radiation at the surface and separated from any continuing water-rock system.<\/p>\n That distinction matters when comparing it with the\u00a0salts, hydrogen and phosphorus detected in material from Enceladus<\/a>. On Saturn\u2019s moon, those chemicals appear to come from a subsurface ocean in contact with rock. On Mercury, water tells us about volatile delivery and preservation, not about a currently habitable environment.<\/p>\n The deposits may be metres thick in some craters, but their depth and volume remain uncertain. The neutron instrument averaged composition over a broad area. Radar is sensitive to material properties and geometry rather than acting as a ruler inserted through the deposit. Estimates ranging from several metres to several tens of metres depend on models and on which craters are included.<\/p>\n Shadow is necessary, but it does not guarantee ice<\/p>\n There is a useful complication in the maps: about half of the permanently shadowed area at each pole is not radar-bright.<\/p>\n A\u00a02018 study of Mercury\u2019s south pole<\/a>\u00a0found the same broad pattern seen in the north. Radar-bright deposits are strongly associated with permanent shadow, yet many shadowed places that appear thermally suitable have no detectable radar-bright material.<\/p>\n Several explanations remain possible. Some cold traps may never have received much water. Ice may have been lost during warmer episodes, mixed with regolith until its radar signature weakened, or buried too deeply for the observations to detect. Small patches can also fall below the resolution of a particular instrument.<\/p>\n This is the same kind of selection problem that appears elsewhere in planetary science. In\u00a0my article on the search for life beneath Mars<\/a>, depth could provide protection while also making the most relevant material difficult to sample. On Mercury, the shelter comes from crater geometry, but the observational problem is similar. The places most capable of preserving a volatile are also the places cameras cannot illuminate directly.<\/p>\n Permanent shadow creates a cold trap. It does not supply the water, date its arrival or ensure that enough accumulated to be visible.<\/p>\n The ice may be younger than the craters holding it<\/p>\n It is tempting to call Mercury\u2019s ice ancient, and on a human scale it almost certainly is. On a geological scale, its age is still an active question.<\/p>\n Water could have come from comet and asteroid impacts, micrometeorites, gases released from Mercury\u2019s interior, or chemistry involving hydrogen carried by the solar wind. Molecules released elsewhere on the surface would make a series of hops through Mercury\u2019s exosphere. Most would be destroyed by sunlight or lost to space, but a small fraction could eventually land in a polar cold trap.<\/p>\n The relatively pure radar signature of some deposits and the associated dark material have led researchers to consider episodic delivery by a volatile-rich impact. A\u00a02020 crater-counting study<\/a>\u00a0argued that exposed ice surfaces may have been emplaced within roughly the past 150 million years. That is young compared with Mercury\u2019s 4.5-billion-year history, although the estimate depends on how impact craters in dark, poorly illuminated terrain are identified and interpreted.<\/p>\n