Triton 495
the interaction of atmosphere and surface. [SeeIO; MARS:
Atmosphere: History and Surface Interaction.]
As described in Section 3, spectroscopic evidence prior
to theVoyager 2encounter indicated that nitrogen existed
on Triton in condensed form.Voyagershowed Triton to be
much smaller, brighter, and colder than had been guessed.
Surface temperatures could be inferred from the visible re-
flectivity as well as measured directly by the Infrared Inter-
ferometer Spectrometer (IRIS). Occultations (passage of
the spacecraft or a star behind Triton) observed by the Ul-
traviolet Spectrometer (UVS) and Radio Science Subsystem
(RSS) probed different parts of the atmosphere, revealing
its temperature and density, from which pressure and com-
position could be deduced. These investigations revealed a
consistent picture of a surface and lowermost atmosphere
at about 38 K. The pressure at the surface was only 14 mi-
crobars, indicating that the gas was in equilibrium with solid
nitrogen at the same temperature. The thermal structure
of the lower atmosphere is not well constrained, but the
temperature probably reaches a minimum at about 8 km
height, above which it increases to about 100 K in the up-
per atmosphere because of heat deposited from space and
conducted downwards. In meteorological parlance, Triton’s
thermosphere directly overlays its troposphere.
TheVoyagerimages and occultation data revealed a va-
riety of condensates in the lower atmosphere. Most of the
atmosphere contains a diffuse haze that can be seen against
the background of space at Triton’s limbs, and which prob-
ably consists of hydrocarbons and nitriles produced by the
action of sunlight on trace gases such as methane. Discrete
clouds were also seen at the limbs and against the unlit part
of the satellite beyond the terminator, where they formed
east-west trending “crescent streaks” roughly 10 km wide,
a few hundred kilometers long, and 1 to 3 km above the
surface. At the limbs, clouds could be distinguished from
haze by being optically thicker and localized both in height
(10 km or less) and in horizontal extent (patchy, and mainly
concentrated at mid to high southern latitudes, where they
cover a third of the limb). The sharper upper boundary to
the clouds suggests that they consist of condensed nitrogen
rather than involatile solids like the haze.
The crescent streaks provide clues to atmospheric mo-
tion by their east-west orientation and the apparent east-
ward motion of the largest, highest cloud seen. Further
clues come from markings on the surface. Over 100 dark
“streaks” were seen in the southern hemisphere, mainly be-
tween latitudes of 15◦and 45◦S. The streaks range from 4
to over 100 km in length, and many are fan-shaped. The
vast majority extend to the northeast from their narrow
end (presumably the origin point); a smaller number are
directed westward. These streaks are extremely similar to
“wind tails” that are common on Mars and are seen on the
Earth and Venus as well. On these other bodies, wind tails
are created by deposition (or sometimes erosion) of loose
material by localized eddies downwind of topographic fea-
tures. It was initially difficult to understand how wind tails
could form on Triton, however, because the atmosphere is
so thin that even the slightest tendency for dust grains to
stick to one another would prevent their being lifted by the
wind.
The interpretation of the surface streaks as wind-created
was nevertheless strengthened by the discovery, shortly af-
ter closest encounter, that some of the streak-like features
were actually atmospheric phenomena. Stereoscopic view-
ing of images obtained from varying angles asVoyager 2
passed by Triton (Fig. 13) revealed that, although the ma-
jority of the streaks were on the surface (or at least too low
to measure their altitude, less than 1 km), at least two had
an altitude of roughly 8 km. These features were subse-
quently named Mahilani Plume (48◦S2◦E, with a very nar-
row, straight cloud 90–150 km long) and Hili Plume (57◦S
28 ◦E, actually a cluster of several plumes with broadly ta-
pering clouds up to 100 km long). Thus, it is clear that winds
on Tritondotransport suspended material, but the question
ishowthe material becomes suspended.
The plumes were entirely unexpected, and explaining
their vigorous activity became a major focus of research as
described below. What is clearest is that they complete a
coherent picture of winds on Triton at the time of the en-
counter. Unlike most surface streaks, both plume clouds
extend westward from their apparent sources (the plumes
proper—narrow, possibly unresolved vertical columns link-
ing the horizontal plume clouds with the surface). Images
of Mahilani appear to show kilometer-sized “clumps” within
the cloud moving westward at 10–20 m sec–^1 , and elonga-
tion of the cloud from 90 to 150 km at a similar speed. Thus,
putting all the descriptions above together (crescent streak
clouds, dark surface streaks, and plume tails), the wind is
northeast nearest the surface, eastward at intermediate al-
titudes, and westward at 8 km, the top of the troposphere.
This is precisely the circulation pattern predicted at the
time of encounter, the height of summer in the southern
hemisphere. Heating by sunlight is presently causing solid
nitrogen in the south to sublimate (evaporate); meanwhile
in the colder north, the atmosphere is precipitating. Be-
cause of the rotation of Triton once every 5.877 days, how-
ever, the wind does not blow directly from south to north to
make up the difference. Instead, gas is transported north-
ward only in a thin skin of atmosphere near the surface
(the Ekman layer) in which the flow is northeastward. The
atmosphere above the 1-km-thick Ekman layer circulates
from west to east. The westward flow at the altitude of the
plumes can be explained if Triton’s atmosphere is slightly
warmer over the equator than at the south pole (perhaps be-
cause the equator is darker), in which case the temperature
gradient will drive a thermal wind that causes the eastward
flow to weaken and eventually change to westward flow with
increasing altitude.
Basic properties of the plumes can be inferred from the
images. The plume clouds do not settle out visibly (no more