496 Encyclopedia of the Solar System
(a)(b)FIGURE 13 (a)Voyager 2image of the southern polar region of
Triton in which geyser-like eruptions were discovered. Here
plumes are viewed obliquely with Hili (H) and Mahilani (M)
plumes marked. (b) Highly magnified images of Mahilani plume
on Triton, taken from increasingly oblique angles and at
increasing resolution (top to bottom). The images have been
projected onto a spherical surface with a viewing geometry
similar to that at the top. The increasing parallax from top to
bottom makes the plume “stem” appear to grow taller. (Courtesy
of NASA/Alfred McEwen, University of Arizona.)
than the∼1-km resolution of the best images) over their
length, so the suspended particles must be smaller than
about 5 mm. From this particle size and the width and con-
trast of the clouds—about 5% darker when seen against
Triton—one can further infer the amount of solids: about
10 kg sec–^1 must be discharged if the material is dark or
twice as much if it is bright. (Bright material in a cloud would
appear relatively dark against Triton’s very bright surface,
though not as dark as intrinsically dark material. However,
bright particles deposited from such a cloud would not show
up as a dark streak on the surface.) The cloud moves hori-
zontally at the wind speed, 10–20 m sec–^1 , but the vertical
velocity in the plume must be significantly faster because
the plumes are not blown visibly askew by the wind. The
columns may be just barely resolved in the best images.
Thus the plumes may be 2 km across or perhaps smaller. The
source area must have similar (or smaller) dimensions. Little
or no structure is visible in the columns, though a “sheath” of
descending material around the plume has been described
by some authors. The active lifetime of the plumes can be
estimated at a few Earth years: shorter, andVoyagerwould
have been unlikely to see any plumes active; longer, and
active plumes should have been more numerous compared
with surface streaks.7.2 Plume Models
Numerous attempts have been made to model the plumes
in order to answer the questions of where the particulates,
the gas suspending them, and the energy to drive the gas
flow originate. Most models have taken their cue from the
presence of the active plumes (and surface streaks) at mid
to high southern latitudes at a season when the sun was
almost directly overhead (Fig. 14 ), and assumed that the
plumes are somehow solar powered. It is also possible, how-
ever, that Triton’s internal heat drives the plumes and that
their location is determined not by the sun but by a local
enhancement of this heat source (i.e., by cryovolcanic activ-
ity) or by the thickness of the nitrogen “cap,” the equivalent
area of the northern hemisphere being hidden in darkness
during the encounter.
It is conceivable that the plumes are purely an atmo-
spheric phenomenon. One early suggestion was that the
plumes are dust devils, localized regions of spinning and
ascending hot atmosphere formed above patches on the
surface that are bare of N 2 frost and that can therefore be
heated by the sun to higher temperatures than their frosty
surroundings. Tritonian dust devils would, however, have
difficulty picking up dust from the surface and becoming
visible, simply because their winds are not strong enough.
If the hot areas on the ground were not only nitrogen-free
but contained methane frost, though, they would give off
clouds of methane gas. Being lighter than nitrogen, this
methane would ascend, and might partially recondense in
the atmosphere, making the rising plume visible. Falling