The last place in the solar system you’d expect to find ice (except maybe on the Sun, duh) is Mercury. Rocky, barren, airless, and very, very hot, Mercury doesn’t sound like the ideal location for storing vast quantities of frozen water.
But in the 1990s evidence started coming in that perhaps Mercury was holding a surprise. At its north pole are deep craters, and because of their high latitude, the low Sun never reaches the crater floors. They’re permanently dark, and because of that they’re very cold. Cold enough to hold on to any water that might have found its way there (presumably through water-bearing asteroid and comet impacts).
The first evidence was from radar observations; those craters were found to be very radar-reflective, which suggested ice, though other materials were possible. But over the years more clues arrived, and when the MESSENGER spacecraft began orbiting the tiny world, the idea of polar water got kick-started. Neutrons were reflected from the crater floors, which indicated the presence of hydrogen (water molecules have two hydrogen atoms each and are very good at reflecting incoming neutrons). MESSENGER has an infrared laser altimeter on board (it uses timing of pulses of light to measure its height off the surface and get topological data), and the craters were again found to be very reflective, which is consistent with ice.
And now we have further, very striking data: Pictures taken of the floor of the crater Prokofiev* show that some of the surface itself is a bit brighter, a bit shinier, than surrounding material. Not only that, but the brighter regions correspond extremely well with what has been found before.
The picture here shows the data. The upper left (A) is from radar observations; the blue circle is the crater rim, the red region is where it’s permanently shadowed—the Sun never shines there—and the yellow is where the radar reflections were brighter than normal. The bottom left (B) shows where the laser altimeter found unusually bright material. On the right (C and D) are the images taken by MESSENGER’s visible light camera. They are the same area and have the same orientation but were taken when the Sun was shining from different directions. The brighter landscape there is clearly visible on the right, and as you can see matches the other observations right on the nose.
The scientists found similar results in other craters even farther north on Mercury (Prokofiev is about 5° south of the north pole, and is 112 kilometers, or 70 miles, across). The amount of ice estimated to be trapped in the floors of these craters is 10 billion to one trillion tons—a huge amount. As the paper points out, that’s about the volume of Lake Ontario.
Personally, I find this to be pretty convincing. It's not a 100 percent lock, but the evidence is getting to be pretty hard to deny.
The ice is likely to be young, too. Impacts, ultraviolet light breaking down the molecules, and other weathering could darken, bury, or eradicate the ice on a timescale of tens or hundreds of millions of years, so it’s likely this deposit hasn’t been around since the early solar system (astronomers define "young" differently than normal humans).
In practical terms, I have a hard time seeing us sending folks to Mercury, setting up a base at its poles, and taking long hot baths using native water any time soon. But this shows that even now, with our huge telescopes, advanced hardware, and robot probes peeking and poking into every corner of the solar system, there’s still a lot to learn about our neighborhood, and a whole lot of surprises waiting to be unwrapped.
We also have similar evidence of water at the Moon’s poles, too, buried under and mixed into the rock at the floors of eternally darkened craters. I don’t have a hard time seeing us going there at all. There could be enough water on the Moon to support a colony for quite some time. That is something I would very dearly love to know more about.
*Craters on Mercury are named after artists: composers, painters, writers, and so on. Sergei Prokofiev was a Russian romantic composer, and one of my favorites; his Fifth Symphony is an astonishing work. It pleases me that such an important discovery has been found in his namesake.