J. L. Bandfield, Dept. Earth and Space Sciences, Univ. Washington, Seattle, WA,,
W. C. Feldman, Planetary Science Inst., Tucson, AZ,, H. H. Kieffer, Celestial Reasonings,
Genoa, NV,
High concentrations of hydrogen detected in Mars’ circumpolar regions are consistent with a high concentration
of water ice in the shallow subsurface [3, 5, 8] and have been modeled as ice beneath a dry particulate layer [4].
The thermal inertia (I) of ice is roughly an order of magnitude greater than typical porous dry particulate regolith
cover, and both the layer thickness and its thermal inertia can be estimated from diurnal and annual temperature
observations. Several studies have taken advantage of these properties to determine the structure of the near-surface
within high latitude localities [11, 1, 2]. Vapor diffusion models predict a water-ice table that generally increases
in depth with decreasing latitude but can vary greatly locally [7, 10]. We derive surface cover thermal inertia and
permafrost depth (Z) at Martian high latitudes from Thermal Emission Spectrometer (TES) data and compare
the results with previously published water ice depths determined from the Mars Odyssey Neutron Spectrometer
(MONS) [4].
Thermal Model:
To compute surface temperatures we use the KRC thermal model, modified from that described in the Appendix of
[6] to have a energy-conserving one-layer spectrally-gray atmosphere. This model allows for customization of a wide
variety of parameters such as changes in subsurface thermophysical properties and atmospheric aerosol properties.
The model was run for two martian years before outputting surface temperatures for the third year.
In a generalized sense, the surface cover thermal inertia controls the offset and the permafrost depth controls the
slope of the seasonal temperature curve. Top layer inertias were allowed to vary from 60-800 J m−2 s−1/2 K−1 , (MKS
units are used for thermal inertia throughout), corresponding to diurnal skin depths of 0.3 to 11 cm. The model
permafrost layer has fixed thermophysical properties with I=2290, but was allowed to vary from 1.15 to 20.3 diurnal
skin depths. The model and fitting routine is sensitive to permafrost at 0.3-6 and 12-220 cm depths for surface cover
thermal inertias of 60 and 800 respectively.
TES observations:
The seasons used for fitting were restricted to summer and early fall seasons (Ls 85.5-220.5 and 265.5-40.5 for the
northern and southern hemispheres respectively). In addition, all surface temperatures below 160K were not use for
fitting because of the proximity to CO2 condensation temperatures. These restrictions as well as the use of only 1-3
hour (H) local time data isolated the model and data from conditions of significant modeling uncertainty, such as low
solar incidence, albedo variation, atmospheric aerosol characteristics or CO2 frost conditions; we avoided ascending
orbit observations at local times of 13-15H. TES data were averaged in bins of 2◦ latitude, 4◦ longitude and 4.5◦ Ls ,
yielding an average of ∼ 90 observations in a bin, with wide variations.
GRS observations and models:
MONS thermal and epithermal neutron count rates are translated into the water-equivalent hydrogen (WEH) abun-
dance of a semi-infinite layer of hydrogen-containing soil having an assumed range of compositions. The top layer
has the same composition as that of the bottom layer, but containing a WEH abundance of 1 wt. % [4]. The burial
depth of the bottom layer is also determined from this model. Because this technique is sensitive to burial mass,
units of g/cm2 are used, convertible to Z using the assumed soil density of 1.5 g/cm3 .
Water ice and solid rock have similar thermal inertia and it is not possible to determine the concentration of water
in the permafrost layer solely from the temperature data used here.
Thermal: The factor of 3 difference in atmospheric dust opacities has a <10% effect on permafrost depth and surface
cover thermal inertia determinations for 2H models but 40% on I and factor of 2 on Z at 14H. Thermal emissivity
changes of 3% , from 0.95 to 0.98, result in 12% change in Z. A similar uncertainty in Z results from the absolute
calibration uncertainty of TES.
Neutron: The depth and fractional abundance of ice depend upon instrument calibration, cosmic ray intensity
variation, the chemistry assumed for the regolith, and the WEH assumed for the upper layer. Modeling done with
upper-layer WEH of 1% and 5% showes that the WEH of the lower layer and its burial depth have a modest positive
dependence upon the upper layer WEH.
Maps of ice depth derived from thermal and neutron observations show strong correlation, although thermal-derived
depths are greater for higher Z values. A general relation between low albedo and surface cover thermal inertia
persists, as derived from thermal-inertia mapping using homogeneous thermal models. There are pronounced regional
deviations from the general correlation of low albedo and greater Z.
Comparison with diffusion models: Maps of ice depth based on modeling with a homogeneous dry-regolith thermal
properties differ from the two-layer models presented here poleward of 70◦ and near the southern rim of the Hellas
basin. These differences are related to the differences in I assumed for the diffusion models, based on [9], and those
derived here. Also, the diffusion model results are for equilibrium conditions, and large regions in the northern high
latitudes have Z values shallow enough that they would be unstable.
THEMIS observations: THEMIS data obtained roughly 2 months apart for a region around Phoenix landing site
were processed in a manner similar to the TES data, with spatial filtering to yield a ∼ 1 km resolution. Average
inertia and depth retrieved from the TES data for the region is 258 and 4.5 cm respectively. Local slopes, surface
cover thermal inertia, and albedo can all have significant effects on the depth of the permafrost. The THEMIS data
has slightly higher values of 283 and 6.2 cm. The THEMIS map displays considerable variation in ice depth, ranging
over the limits of the method from 3 to 30 cm. THEMIS data will be important for placing the Phoenix observations
in the regional and global context.

 [1] J. C. Armstrong, T. N. Titus, and H. H. Kieffer. Evidence for subsurface water ice in Korolev crater, Mars.
     Icarus, 174:360–372, 2005.
 [2] J. L. Bandfield. High-resolution subsurface water-ice distributions on Mars. Nature, 447:64–67, 2007.
 [3] W. V. Boynton et al. Distribution of hydrogen in the near surface of Mars: Evidence for subsurface ice deposits.
     Science, 297:81–85, 2000.
 [4] W. C. Feldman et al. Vertical distribution of hydrogen at high northern latitudes on Mars: The Mars Odyssey
     Neutron Spectrometer. Geophys. Res. Lett., 34:5201–+, 2007.
 [5] W. C. Feldman et al. Global distribution of near-surface hydrogen on Mars. J. Geophys. Res., 109(E18):9006–+,
 [6] H. H. Kieffer et al. Thermal and albedo mapping of Mars during the Viking primary mission. J. Geophys. Res.,
     82:4249–4291, 1977.
 [7] M. T. Mellon, W. C. Feldman, and T. H. Prettyman. The presence and stability of ground ice in the southern
     hemisphere of Mars. Icarus, 169:324–340, 2004.
 [8] I. G. Mitrofanov et al. Soil Water Content on Mars as Estimated from Neutron Measurements by the HEND
     Instrument Onboard the 2001 Mars Odyssey Spacecraft. Solar System Res., 38:253–257, July 2004.
 [9] N. E. Putzig, M. T. Mellon, K. A. Kretke, and R. E. Arvidson. Global thermal inertia and surface properties
     of Mars from the MGS mapping mission. Icarus, 173:325–341, 2005.
[10] N. Schorghofer and O. Aharonson. Stability and exchange of subsurface ice on Mars. J. Geophys. Res.,
     110(E9):5003–+, 2005.
[11] T. N. Titus, H. H. Kieffer, and P. N. Christensen. Exposed water ice discovered near the south pole of Mars.
     Science, 299:1048–1051, 2003.

To top