Onset of Impact-Generated Hydrothermal Systems: Hydrocode Modeling

Pierazzo E., N.A. Artemieva, B.A. Ivanov

3rd Intern. Conf. on Large Meteorite Impacts, 2003, Abst. #4102.
35th LPSC, 2004, Abst. #1352.

For more information contact betty@psi.edu

Mars is the most Earth-like of the solar system's planets, and the first place to look for any sign of present or past extraterrestrial life. Its surface shows many features indicative of the presence of surface and sub-surface water, while impact cratering and volcanism have provided temporary and local surface heat source throughout Mars geologic history.
Impact craters are ubiquitous indicators for the presence of sub-surface water or ice on Mars. In particular, the presence of significant amounts of ground ice or water would cause impact-induced hydrothermal alteration at Martian impact sites. The realization that hydrothermal systems are possible sites for the origin and early evolution of life on Earth has given rise to the hypothesis that hydrothermal systems may have had the same role on Mars. A better understanding of the evolution of material's thermal state underneath impact craters is necessary for more realistic models for the formation of hydrothermal systems on Mars.
We present preliminary results of two-dimensional (2D) and three-dimensional (3D) simulations of impacts on Mars aimed at constraining the initial conditions for modeling the onset and evolution of a hydrothermal system on the red planet. Simulations of the early stages of impact cratering provide an estimate of the amount of shock melting and the pressure-temperature distribution in the target caused by various impacts on the Martian surface. Modeling of the late stage of crater collapse is necessary to characterize the final thermal state of the target, including crater uplift, and the distribution of the heated target material (including the melt pool) and hot ejecta around the crater.

EARLY STAGE:

We use the 3D hydrocode SOVA to model spherical comets and asteroids of various sizes impacting Mars' surface at 15.5 and 8 km/s, respectively. These velocities roughly correspond to median impact velocities for short-period comets and asteroids. Simulations model 90° (vertical), and 45° (most probable impact angle) impacts. A spatial resolution of 20 to 25 cells-per-projectile-radius (cppr) is maintained over a central region around the impact point, followed by regions of progressively lower resolution, extending to about 13 km downrange (5 km uprange), and 15 km below (9 km above) the surface. Over 500,000 Lagrangian tracers are used to mark each computational cell of the target around the impact point (of given volume), and record the maximum shock pressure experienced by the material in each cell. Melt volume estimates are determined by adding up the volume of tracers experiencing shock pressures above a given threshold. We use 46 and 56 GPa as threshold pressures for incipient and complete melting of pure granite.

Table 1: Shock melting associated with various 3D impact simulations on Mars.
Type Dpr (km) vimp (km/s) Impact Angle Drim (km) Vm (km3)
Comet 2 15.5 90° 26-34 31.6
Asteroid 2 8 90° 26-34 4.1
Comet 2.32 15.5 45° 26-34 32.5
Asteroid 2.32 8 45° 26-34 3.7
Comet 8 15.5 45° 69-90 444
Asteroid 8 8 45° 69-90 91

Table 1 shows melt volumes associated with impacts producing final craters (rim to rim) of about 33 and 80 km, according to Pi-scaling laws, for vertical and/or 45° impacts of comets and asteroids. These results indicate that impact velocity plays an important role in the total volume of melt produced in the impact. In particular, because of their larger impact velocities cometary imapcts are much more efficient in creating larger melt pools whose longer cooling time may contribute to a longer duration of the impact-generated hydrothermal system. The effect appears to be stronger for the smaller craters. While the initial simulations use a single material (granite) target, a more realistic target is one in which the basic crystalline crust contains a component of ice and/or liquid water. Initial simulations model a 2-km asteroid impacting at 10 km/s and 45°. Four different targets have been modeled: 1) Dry granite target; 2) target with 10% ice distributed in regularly spaced horizontal layers (1 every 10 rows are made of water); 3) target with 10% ice distributed in regularly spaced vertical layers (1 every 10 columns are made of water); 4) target with 10% water mixed throughout (every cell contains 10% of ice).

The figure shows volume of target material compressed above a given shock-pressure normalized by projectile volume. It is evident that the presence of water ice in the target does not affect the overall shock decay for shock pressures above about 20 GPa. Below 20 GPa the various cases differ by up to 15%, but this may be an artificial effect, due to the cutoff in tracers distribution away from the impact point. On the other hand, mixed materials in the target may affect shock propagation and consequently the final estimate of melting. If the target is a mixture of materials with very different impedances, as is the case for rocks and water, the shock wave propagation may be affected by the interaction of the original shock wave with shocks reflected at material boundaries. In shock events it is only the first shock that heats the material, while secondary reflected shocks increase the final shock pressure, with a minor contribution to the temperature. As a result, single- and multiple-shocked materials experiencing the same peak shock pressure will reach substantially different thermodynamic states, with the multiple-shocked material having lower shock temperature and shock entropy. Therefore, for mixtures of materials with very different impedances the maximum shock pressure alone may not be the best criterion for estimating melting and vaporization.
We are planning to carry out a more detailed series of simulations to investigate the pressure-temperature conditions in compression and decompression, to estimate melting in mixed targets.

LATE STAGE:

The final temperature field around an impact crater depends both on the shock compression/decompression cycle and on friction heating due to plastic deformation. As the geothermal heat flow gradient is responsible for an increase in temperature downward, material uplifted from below during the formation of the central peak/peak ring in complex craters is at higher temperatures that surrounding material, thus providing a further source of heat. For a complete picture of the thermal field underneath an impact crater it is thus necessary to follow the entire crater-forming event, from impact to the final crater.

We use the 2D hydrocode SALE2B to model crater collapse and the formation of a final crater around 30 km in diameter (rim-to-rim), using the 2D hydrocode SALE2B, which has already been used to model the formation and thermal evolution of several terrestrial impact structures of various sizes. The simulations covered both asteroid (granite), and cometary impacts (ice), at 8 and 15.5 km/s. The target is modeled as crystalline granite, with a thermal gradient of 13 K/km (Babeyko & Zharkov -PEPI 117:421-435, 2000- assume 14 K/km as an "average" gradient estimate for Mars) and a starting surface temperature of 298K (typical cited Mars surface temperature is around 220K), which are characteristic terrestrial conditions, and a typical Martian gravity of 3.72 m/s2.
Assuming that the pore water has the same temperature as the surrounding material, we use the pure water/ice phase diagram to identify the final phase of water underneath the crater. The figure shows that the resulting thermal fields underneath same size craters created by asteroid and comet impacts are similar: the combination of shock/plastic heating and the structural uplift of initially deeper strata create a water-bearing zone at depths where water is in the liquid stability field. The main difference is that there is a larger volume of hot rocks in the central peak of the comet produced crater, which is a natural result of larger initial shocks near the impact point due to the larger comet impact velocity. In the central uplift the high temperatures cause water to evaporate (steam-driven circulation). The liquid and vapor water stability zone characterizes the onset of a hydrothermal circulation cell. The figure suggests that for a mid-sized crater (rim diameter around 30 km) the hydrothermal circulation is probably restrictred to a "column" limited to about one-half of the final crater.
These results are to be considered preliminary, as much testing is still necessary to characterize the best crater collapse model parameters for Martian conditions.