Triton 499
on Mars. In this case the polar cap material and geyser gas
is not N 2 , but CO 2.
7.3 Polar Cap and Climate
We turn now from the plumes to a consideration of how
Triton’s surface frosts and atmosphere change over time.
Here, too, theVoyagerimages yielded a surprise: at the
height of southern hemisphere summer (Fig. 4), most of
the southern hemisphere was covered with a bright deposit
(a polar cap), but the visible portion of the northern, winter
hemisphere was darker. Models of the redistribution of N 2
frost with the seasons can be constructed with varying de-
grees of complexity, but a fundamental expectation is that
the summer hemisphere should have less of a polar cap than
the winter one!
The basic physics of seasonal frost-distribution models
is as follows.
1.The whole atmosphere and all frosted areas are at very
nearly the same temperature. If a frosted area were colder,
more nitrogen would condense there and release of latent
heat would raise the temperature. Conversely, a warm frost
area would be cooled by sublimation. Winds would quickly
even out the atmospheric pressure and temperature.
2.At this fixed temperature, sublimation occurs where
frosts are exposed to the sun and condensation where
the average input of solar energy is less. Sublima-
tion/condensation rates can be calculated from the amount
of sunlight absorbed at each point on Triton.
3.Bare (unfrosted) areas can be warmer than the atmo-
sphere and frosts (if they are dark and/or well exposed to the
sun) but they cannot be colder, or frost would immediately
condense on them.
Using the albedo of the surface as measured byVoy-
ager, models indicate that frost in most of the southern
hemisphere is currently subliming, thinning the surface de-
posits. Nitrogen is presumably being deposited in the north-
ern hemisphere and in a few of the brightest areas of the
south where little sunlight is absorbed. Stellar occultations
sinceVoyagerhave shown that Triton’s surface pressure
(and thus atmospheric mass) has measurably increased, to
around 19μbar! By inference the surface temperature of
the nitrogen ice, which controls the atmospheric pressure,
has also increased by 1–2 K. But what about the long run?
By assuming that frost has some given albedo and that the
surface underneath has some other albedo, one can model
the redistribution of nitrogen over long periods. A layer of
nitrogen frost about a meter thick is moved back and forth
as the sun shines on one hemisphere and the other, and the
pressure and temperature of the atmosphere change as well.
Notably, such models predict that all nitrogen deposited in
the southern hemisphere the last time it was winter there
would have resublimated beforeVoyagerarrived. Corre-
spondingly, the northern hemisphere should be extensively
frosted.
How can these predictions be reconciled with observa-
tion? The frost might deposit mainly in shadows and on
north-facing slopes whereVoyagercould not see it, or it
could be glassy and transparent, hence invisible. There is
some evidence for the last possibility, from laboratory ob-
servations of condensing nitrogen, calculations of the rate at
which loose frost grains would merge or anneal into a dense,
transparent layer, and even from observations of the light-
scattering properties of Triton’s equator. Such suggestions
would each explain the dark, apparently frost-free northern
hemisphere, but the bright “cap” in the south must be ex-
plained as well. Perhaps it is a much thicker deposit of nitro-
gen that never completely sublimes away (this is certainly
the impression one gets geologically). Although nitrogen
frost may be very transparent when first annealed, chang-
ing temperatures will make the residual cap expand and
contract, fracturing it and making it appear bright. Thus,
we are led to the idea of a clear, uncracked (i.e., gas-tight)
seasonal frost layer over a thick, fractured permanent cap:
precisely the kind of layering hypothesized above to explain
the plumes as solar-powered geysers.
What controls the size of the residual cap, and why is
one not seen in the north? A good candidate is solid-state
creep, or flow, of the thick nitrogen deposit, similar to the
flow of glaciers and spreading of polar caps on the Earth and
Mars. Models based on terrestrial polar caps, combined
with estimates of the rate at which solid nitrogen would
flow, suggest that the permanent cap is about a kilometer
thick at the center. Cap spreading also prevents the even-
tual disappearance of the seasonal frosts predicted by the
models discussed above. Because the pole always receives
less sunlight than the edges of the seasonal frost deposits,
more frost will be deposited at the pole than at the edges,
maintaining the cap. There may be a northern as well as a
southern permanent cap. If this northern cap extends less
than 45◦from the pole, it would lie in the dark portion of
Triton unseen byVoyager. The southern permanent cap
might be larger because of hemispheric differences in the
heat released from Triton’s interior, or it might also extend
only 45◦, in which case the bright deposits extending al-
most to the equator have still to be explained. Some of this
bright material may be nitrogen “snow” that condenses in
the atmosphere into grains that are too big to anneal on
a seasonal time scale into a transparent layer. It should be
apparent from this discussion that, as with the plumes, we
seem to have many pieces of the puzzle of the polar caps
(and perhaps a few spurious pieces of unrelated puzzles),
but they have yet to be assembled into a final picture of
Triton’s surface-atmosphere interaction.
Additional clues to the behavior of volatiles on Triton are
presently being gathered from Earth-based spectroscopic
measurements, and by the occultation of stars by Triton.
As noted above, Triton’s atmosphere is changing, becoming