A winning proposal for the Innovative Research Program, 2005:

Sea ice on Mars? Earth-Mars Satellite Altimetry Comparison and Theoretical Modeling

Investigators: Todd E. Arbetter (CIRES/NSIDC), Ted A. Scambos (CIRES/NSIDC)
CIRES Research Theme: Advanced Observation and Modeling Systems

Objectives: Investigate the possibility of relict frozen sea ice in Elysium Planitia, Mars, by comparison of Martian and Terran satellite imagery and satellite elevation data. Examine the likely characteristics of sea ice on Mars using appropriately modified dynamic sea ice models.

Fig. 1

Background and importance

Patterns in recent imagery from the High Resolution Stereo Camera (HSRC) aboard the Mars Express spacecraft suggests the existence of a frozen sea of water ice in the Elysium Planitia region of Mars (Murray et al., Nature, March 17, 2005). The features, estimated to be approximately 5 million years old, cover an area 800 km by 900 km, similar in size to the North Sea. Estimated past water depth (from basin topography) is 45 meters. The features are claimed to be inconsistent with solidified basalt lava flows observed elsewhere on Mars, although previous studies suggested this. Size and scale of the fracture and floe patterns suggests (to Murray et al.) a layer of ice once floated above a liquid sea (Figure 1, right). The liquid drained or sublimated, leaving a dust-covered, relict, sea-ice-like pattern.

A comparison of images from Earth and Mars seems to superficially support the contention by Murray et al. (Figure 2). It should be noted that solid basalt does not float on liquid basalt magma, and so directly analogous igneous processes are not possible. Closer inspection of the rubble field around one of the craters (Figure 2) shows it is analogous to the flow of plates around icebreakers (Figure 3).

Fig. 2

Figure 2: HSRC image of 'fractured plate'-like features in Elysium Planitia (left). Antarctic pack ice (right). Murray et al. (2005) hypothesize that the Martian features are remnants of an ice-covered sea. Note the 'puzzle-fit' shapes of the large floe and 'shelf' along the right edge of the Martian image.

Fig. 3

Figure 3: Closeup of a crater in Elysium Planitia. The diameter is approximately 1.5 km. The existence of a rubble field to the right of the crater and a lead-like feature to the left suggest that the crater may have impeded a floe of ice traveling from right to left. Features within the crater suggest a floe which grounded within the crater as the water level subsized. (Murray et al. 2005).

Why is this important? The HSRC imagery represents the most direct evidence to date of the recent existence of a large body of liquid water on the Martian surface. The presence of liquid water is essential to speculation that life may have existed, or may exist, on Mars.

What makes this innovative? We will use a modified sea ice model as the basis for a Mars-analog ice-covered ocean, and explore what kind of large-scale sea ice fracture and drift patterns might be expected on Mars using idealized experiments. We will compare Mars and Terran laser elevation profiles (from MOLA and GLAS) in search of sea-ice structures that may be used to resolve the debate.

Research plan

Observationally, we will gather laser altimetry data from both Arctic and Antarctic sea ice (using ICESat's GLAS data) and the Elysium Planitia region using MOLA. ICESat data is archived at NSIDC, and readily available. MOLA data is at http://pds-geosciences.wustl.edu/missions/mgs/mola.html, complete with tools for plotting profiles and information on the use of the data. Using satellite images and the profiles, we will compare structures and scales of the two surface types (Martian sea-ice-analog(?) and Antarctic sea ice).

A viscous-plastic rheology sea ice model with a multiple-category distribution resolving relative fractions of level and ridged ice in a distribution of thicknesses (Flato and Hibler 1995) will serve as the basis for the model experiments. The dynamics of this model are governed by a momentum balance which includes wind and ocean drag terms as well as an internal ice strength parameterization. The Coriolis term will be modified to be appropriate for the Martian planetary radius and rotation rate, with the assumption that the southern boundary lies along the equator. Because of the lower Maritan gravity, less energy will be required for ridging (Rothrock 1975, Hibler 1980, Flato and Hibler 1995), so the strength and deformation of sea ice will be altered. A ~160x180 domain of 5km resolution, with shoreline and bathymetery features approximated from the Elysium Planitia morphology, will serve as the testbed for initial model experiments. These will include idealized wind forcing with uniform and sinusoidally varying wind stresses including the possibility of a shear zone across the ocean (for example, westerlies in the northern half of the ocean, easterlies in the southern half). The growth rate of the Martian sea ice will be determined by an estimated flux balance, assuming a constant air temperature across the model domain. The dominant terms in the flux balance equation will be downwelling shortwave radiation, radiative cooling, and conduction of heat through the ice. Because of the thin Martian atmosphere, we will assume that turbulent heat fluxes and downwelling longwave radiation are negligible.

An important difference between the two planets in this analog comparison is the lower Martian surface air temperature and atmospheric density. In separate, simplified models, we will investigate the basic processes of Martian sea ice growth and thickening, by using appropriately modified thermodynamic algorithms. Martian water bodies are likely very saline, with sulfate salts (as recent results from the Spirit and Opportunity rover imply); these will change the freezing point and growth rate, as well as growth characteristics. For example, in a nearly-saturated salt water body at low to extremely low temperatures (e.g. -20C to -80C, typical of Mars), the process of ice growth and brine rejection would lead to evaporite precipitation by freezing.

Expected outcome and impact: The detailed comparison of MOLA and IceSat imagery, as well as model results, will offer more evidence to support or contradict the hypotheses of Murray et al. The results from the model experiments will explore possible behavior of a Martian ice-covered sea. Further, the IRP grant will serve as an important 'seed money' project to assist in the development of more-detailed, larger grants to NSIDC from NASA given its new Moon-Mars exploration focus.