A global colour mosaic of the part of Triton imaged by Voyager 2 in 1989.  The south polar cap is at the bottom and the cantaloupe terrain north of it. (Source: NASA/JPL/USGS)

Triton – a mysterious encounter

A global colour mosaic of the part of Triton imaged by Voyager 2 in 1989. The south polar cap is at the bottom and the cantaloupe terrain north of it. (Source: NASA/JPL/USGS)

In 1989, Voyager 2 approached Triton, by far Neptune’s largest moon. We already knew it was very cold, near the freezing point of nitrogen, and there were great hopes of finding a world such as Titan turned out to be, but with lakes and rivers of nitrogen rather than methane. That didn’t turn out to be the case: Triton has a very high albedo, reflecting most of the sun’s light away from it, and that leads to a surface temperature of just 38K – cold enough for just a thin nitrogen atmosphere and a nitrogen, CO2 and water ice surface.

But what a surface! Strange and mysterious textures, an array of cryovolcanic and tectonic features, even dark plumes rising off from its south polar region to heights of 8km. Frozen it may be, but Triton’s an active world, with a surface that may be as little as a few million years on the basis of impact crater counts1.

Neptune’s Captive

Triton is unique amongst full-sized moons in the Solar System in that it orbits Neptune in the opposite direction to the planet’s rotation – a retrograde orbit. That wouldn’t occur if it accreted from the original planet-circling debris disk, so it was probably captured by the gas giant after its formation. The leading theory is that it formed in the Kuiper Belt in much the same way as Pluto, but was part of a binary with another (possibly much smaller) body. During a close encounter between that binary and Neptune, the planet took the part of the other body in the binary, and that other body was ejected.2

All this action in its young life gave Triton a lot of energy: it would initially have had a very eccentric orbit and the tidal stresses caused as Neptune circularised that orbit and changed its rate of spin led to a huge amount of heating. Added to that, since it’s a body like Pluto which formed from the initial solar nebula and not the leftovers from gas giant accretion, it has a higher rock percentage (c.65-85%) than any of the other moons we’ve seen. The rock contains radiogenic isotopes which provide more heat. It’s thought Triton melted completely after capture over a period of up to a billion years3, leaving it a fully differentiated body.

Its orbit is circularised now so it’s not getting intense tidal heating, but there’s probably heat left over from the processes which got it to this state and it still has heatflow from the radioactive isotopes. It may even still have a liquid subsurface water-ammonia ocean, depending how eccentric its original orbit was4.

Nitrogen-Driven Geysers

Dark streaks on Triton's south polar region, the sites of 8km-high geysers (Source: NASA/JPL)

The detection by Voyager II of dark plumes of material coming off the surface of Triton made it one of the few bodies in the Solar System which we known to have present-day eruptive activity. There are also many dark streaks from earlier eruptions, the quantity of which suggest each plume has a lifetime for 1-5 Earth years. The favoured theory for their formation is that the heat of the sun passes through nitrogen ice but gets trapped there as in a greenhouse. Because the surface of Triton is close to the freezing point of nitrogen, it would only take an increase of 4K to make the nitrogen sublime to gas and cause a build-up of gas pressure. At a certain critical point, it will shoot up through the surface like a geyser, lofting up dark particles which may have been incorporated in the crust by the photodissociation of methane.

This theory fits the evidence pretty well: the plumes concentrate under the point on the surface which is closest to the sun, and the physics of the process seem to add up5. They’re usually cited as evidence for cryovolcanism on Triton, but this if this is how they happen, I don’t think they qualify: they’re powered by external solar heat, not the heat of the moon, so they’re really an exogenic process.

There’s another theory, however, that they occur at the subsolar point because that’s the area with the most volatiles to power the geysers, rather than because of the sun. In that model, the heating required to melt the nitrogen would be internal, from the decay of radioactive nucleides in the rocky part of the planet, in which case the analogy of geysers like Old Faithful in the USA is a closer one and it’s more reasonable to refer to them as cryovolcanic.

Smooth Cryoflows

Evidence of cryovolcanism: a chain of pits and mounds in the centre of the image are probably the vents from which the smooth material surrounding them erupted. (Source: NASA/JPL/USRA/LPI)

We’re on safer ground with the many smooth flow deposits and associated pits and depressions found on Triton. There are huge smooth-floored depressions which look like frozen lava lakes, depressions surrounded by smooth materials to distances of up to 200km and chains of pits which are very similar to collapsed lava tubes and fissure vents on Earth6. Other possible cryovolcanic features are dark patches which look like flows of a different, possibly more viscous, composition, and surface-softening deposits which could be cryoclastic (similar to ashfalls on Earth).

The composition of these flows isn’t entirely clear. Again we’d need a substance that is buoyant enough to rise above a low density icy shell. Aqueous ammonia may again be the answer, with their explosivity powered by nitrogen as well as possible methane from methane clathrates7.

A criss-crossed cantaloupe

The dimples and double-ridges making up the cantaloupe terrain. The image is 220km across. (Source: NASA, JPL)

Titan’s oldest and most distinctive surface is known as cantaloupe terrain. This is a huge area in the western hemisphere covered by regularly-spaced dimples, each about 30-40km across. Explanations of this terrain range from less dense ices rising through denser layers in diapirs as salt does on Earth, to small explosive volcanic vents8, to a structural cause during the long-lived cooling of the moon9. None of the explanations seem really convincing in light of the scale and regularity of these features.

This terrain is also criss-crossed by troughs and double ridges which are similar to, though wider than, those on Europa. These may have formed by the same process, for example when diurnal tidal stresses caused melting of the icy surface by shear heating along fractures. The extra width can be accounted for by a greater depth of faulting in Triton’s cold, thick lithosphere10. This would have to have occurred some time ago, though, as such stresses aren’t in play now that Titan’s orbit is circularised.

Another sign of tectonics are ‘packets’ of parallel fractures or ridges which look like extensional faults. Though their cause is not clear, they do tend to have preferred orientations which may suggest they formed by global expansion as Neptune’s pull changed Triton’s rate of rotation11.

A note of caution

Intriguing as all these features and terrains are, we need to be cautious in drawing conclusions about Triton. Our only data come from Voyager’s single flyby, which imaged only 40% of the surface, and spectral measurements from here on Earth. This is of such low resolution it could be giving us entirely the wrong impression. Galileo and Cassini have shown just how different icy moons can be on a second, closer viewing – findings from both those missions shattered former theories formed on the basis of Voyager data. It would be great if a mission with those kinds of capabilities could make a return visit to Triton, but, far out and inaccessible at the ends of the Solar System, it may be some time before this moon becomes more than a fascinating mystery.

  1. Schenz & Zahnle (2007) Icarus: 192, 135-149 []
  2. Agnor & Hamilton (2006) Nature: 441, 192-193. []
  3. Boyce (1993) Lunar & Planetary Science Conference XXIV []
  4. Gaeman et al. (2012) Icarus: 220, 339-347 []
  5. Kirk (1990) Lunar & Planetary Science Conference XXI []
  6. Croft (1990a) Lunar & Planetary Conference XXI []
  7. Kargel & Strom (1990) Lunar & Planetary Conference XXI. []
  8. Croft (1990a) Lunar & Planetary Conference XXI []
  9. Boyce (1993) Lunar & Planetary Science Conference XXIV []
  10. Prockter et al. (2005) Geophysical Research Letters, 32, L14202. []
  11. Collins & Schenk (1994) Lunar & Planetary Science Conference XXV []