This page accessed
[TextCounter Fatal Error: Could Not Write to File __klassen_papers_dps95_shtml]
Infrared Imaging Spectroscopy of Martian Clouds and Volatiles
as presented at the 27th Annual Meeting of the Division
for Planetary Sciences of the American Astronomical Society in Mauna Lani,
D. R. Klassen (Univ. of Wyoming)
J. F. Bell (Cornell University)
R. R. Howell (Univ. of Wyoming)
The Data Set
The work here uses a subset of a much larger image set taken by Jim Bell at the
NASA Infrared Telescope Facility using the new NSFCAM 256x256 InSb camera. (More
information on this larger data set can be found in
Jim Bell's presentation.
Note: this is a one-way link so to get back here you will have to use the
'back' function of your Web browser.) The images were taken using the camera's
circular variable filter (CVF) which has a spectral resolution of about 2%. The data
for this work were taken on 1 FEB 95 at an Ls= 54°
(northern spring on Mars).
Recent Hubble images have shown that during the northern spring season Mars has shown
considerable cloud coverage, even at tropical latitudes. This at a time when the
atmospheric temperature should be too high to condense the atmospheric volatiles.
Our spectral image set contains images at wavelengths that include the absorptions bands
of both H2O and CO2 frosts. Using these images we will look for
and track the clouds and ice cap. What makes this set superior to the Hubble images
is that in the infrared we can discriminate between the two types of volatiles and thus
indirectly be able to measure the temperature of the atmosphere.
The images are reduced using standard infrared techniques. They are first linearized,
then sky-subtracted to remove any dark-current, bias-offset and sky flux, and finally
flat-fielded to equalize the gains of each pixel. Analysis then consists of creating
Relative Band Depth (RBD) images. These images are created from the normalized in-band
image ratioed to a normalized continuum image created from a linear interpolation of
two continuum images on either side of the band. This is represented mathematically
The choices for the 1 and 2 continuum and in-band (b) wavelengths were:
These choices were made by applying the RBD formula to published ice spectra. Using
these laboratory ice reflectance spectra, a line is fit between the points being used
as the continuum wavelengths and the spectra are then divided by this line. The
following plots show these band depth spectra:
In these graphs, the circled points are the continuum wavelengths and the lowest point
is the in-band wavelength. As can be seen, these pure volatile band depths are easily
visible, being on the order of a 50% absorption.
The band depth images created are presented here in comparison to images created using
Viking albedo and topography data.
For orientation, north is at top, the bright central region is Arabia, the dark
region to the south east is Syrtis Major, the large dark region in the west is Acidalia
and the bright region in the south (which is seen to be dark, or low, in the topographic
image) is the basin Hellas.
The RBD images were normalized to the bright region Arabia. As a note, the dark/bright
vertical/horizontal bar edge effects and the bright/dark limb edge effects are due to
image registration and are not real effects. In the 3.33µm RBD not much is
seen. Any patterns are only of the order of 2% above or below Arabia. This would be
an indication of either no CO2 frost or at least none relative to Arabia.
It seems plausible that the springtime atmosphere of Mars is not cold enough to condense
CO2 into clouds. However, in the 3.00µm RBD image, one can easily see
the topographically low basin of Hellas as well as the northern lowlands. These differ
from Arabia by 20-30% and would indicate less absorption in these regions. This would
then imply some kind of H2O frost in the darker regions of the RBD, most
likely in the form of clouds.
Again, for orientation, north is at top. Since these images come from the second CVF
scan, Mars has rotated considerably. The bright region Arabia is now near the eastern
limb and the northern dark region Acidalia is near the center of the disk. In the
topographic image the long dark streak in the south is Valles Marinaris and the Tharsis
plateau is just over the western limb. Since Arabia has move toward the limb, to avoid
any limb darkening confusion the bright region Chryse, just south of Acidalia is used
as the normalizing region.
In the 3.33µm image the variations are again only on the order of 3-5% from Chryse.
So this would indicate either a uniform CO2 frost cloud coverage or none at
any detectable levels. The latter is the more reasonable interpretation. However, in
the 3.00µm RBD image patterns with variations on the order of 10% can be seen.
These pattern also appear to follow the topography; one can easily see the two lobed
structure near the western limb in both the RBD and the topographic maps. The darker
areas in the RBD again indicate more absorption than the normalizing region and since
these are over the higher topographic regions this would indicate high, cold
H2O frost clouds.
One interesting point is that the polar ice cap, which should still be very visible at
this season (see the above Hubble images) is missing in both the CO2
and the H2O images. For the CO2 images this has yet to be
explained. However for the H2O images this can be explained. It can be
assumed that the ice at the polar region would be of a fairly coarse, perhaps blocky,
nature. The wavelengths and continua points for the 3.00µm RBD map were chosen to
best show fine H2O frosts and coarser frosts are less visible with these
wavelength choices by about a factor of two.
Our interest lies in absolutely calibrating all the images in the spectral scans using
the stellar image data gathered at the same time and then to make true band depth maps
instead of the relative band depth maps presented here.
We believe, based on this preliminary work, that it will be possible to locate and track
the atmospheric frost absorptions throughout the 1995 opposition. It will be interesting
to see if, during the early spring, the atmosphere was ever cold enough to support
For more information mail to: