Lunar and Planetary Science XXXVI (2005) 2374.pdf MELTING IN MARTIAN SNOWBANKS A. P. Zent, B. Sutter, Ms 245-3 NASA Ames Research Center, Moffett field CA, 94035, [email protected] Introduction Precipitation as snow is an emerging which predicts the temperature and redistribution of paradigm for understanding water flow on Mars, snow. The transport of liquid water and water vapor which gracefully resolves many outstanding are included in the model as necessary elements of uncertainties in climatic and geomorphic the heat balance calculation. The model assumes interpretation [1]. that the Martian surface at the base of the snowbank Snowfall does not require a powerful global is frozen and impermeable. greenhouse to effect global precipitation. It has long The snowbank is allowed to compact under the been assumed that global average temperatures > weight of the overburden. The overburden is 273K are required to sustain liquid water at the augmented in polar winter by the mass of the surface via rainfall and runoff. Unfortunately, the seasonal ice cap (see the seasonal cap discussion). best greenhouse models to date predict global mean The compaction routine is taken from Anderson [9], surface temperatures early in Mars' history that differ and is a linear function of the overburden pressure. little from today's, unless exceptional conditions are As the snow compacts, the finite difference grid on invoked [2]. Snowfall however, can occur at which it is calculated is allowed to compress, so that temperatures < 273K; all that is required is saturation the volume elements continue to correspond with the of the atmosphere. At global temperatures lower than original sample of snow. The density of water vapor 273K, H O would have been injected into the and dry air are invariant during matrix deformation, 2 atmosphere by impacts and volcanic eruptions during and a portion of gas is expelled from the contracting the Noachian [3.4], and by obliquity-driven climate volume, which is taken into account when defining oscillations more recently [5,6]. Snow cover can fluxes for these constituents. accumulate for a considerable period, and be available for melting during local spring and summer, The snow energy balance equation accounts for the unless sublimation rates are sufficient to remove the energy associated with the mass flux of liquid water, entire snowpack. water vapor, thermal conduction, and radiative flux. We decided to explore the physics that controls the It is assumed that only the short-wavelength solar melting of snow in the high-latitude regions of Mars radiation penetrates the top node, and a maximum to understand the frequency and drainage of depth for solar penetration of 30 cm is assumed. The snowmelt in the high martian latitudes. extinction coefficient is currently assumed to be that The surface mass and energy fluxes of the snow of pure snow, [9]. Although the introduction of model are calculated by mass and energy balance snow/dust mixtures will be implemented in the next with a one-dimensional radiative-convective version of the model. boundary layer model [7]. The model predicts Latent heat changes, either due to movement of liquid atmospheric temperatures caused by radiative (solar water or phase change within an element are and infrared) and non-radiative (convection, included. The apparent heat capacity method [10] is turbulence, etc.) effects. Variations in surface winds caused by frictional mixing in the planetary boundary layer are also computed, in order to calculate turbulent energy fluxes to the snow interface. In the simulations described here, the optical depth of H O 2 ice is assumed zero, although the code exists to explore these effects. The code calculates the solar absorption by atmospheric CO . It considers separately infrared 2 absorption by atmospheric CO inside the strong 15 2 µm band. The model adjusts temperatures due to convection, and pases to the snowbank model the temperature of the air just above the surface, as well as the direct and diffuse solar insolation, and the downwelling IR radiative flux. Fig. 1. The snowbank model solves energy and We use a one dimensional mass and energy balance mass balance for the snow and overlying CO cap. model adapted from sntherm.89 [8]. 2 Lunar and Planetary Science XXXVI (2005) 2374.pdf used, where the total enthalpy change is expressed in [6] Kieffer and Zent, 1992 terms of temperature through the definition of an [7] (Haberle et al., 1993) apparent specific heat. [8](Jordan, 1991), The surface energy balance of the snowbank is [9] Anderson (1976) composed of the turbulent fluxes of sensible and [10] Albert (1983) latent heat, and the short and long wavelength [11] Kieffer and Titus (2000). components of the radiation. The turbulent exchange fluxes depend on surface roughness, wind speed and the atmpspheric gradients of temperature and humidity. The radiation term is calculated by the coupled 1 Snowbank Sublimation - 60 N dimensional radiative-convective Mars atmosphere 0.6 code of [7]. The albedo of snow is taken as 0.34, as a Obliquity = 35 basline the average value of the Martian polar caps. 0.5 Relatively fresh, pure terrestrial snow has an albedo m) otAshufte r asfarelaoc eteuix tnuctdrode oe0msl.s 7e s8pt o oc[ al8tehn]we ,b aaeprn odedi xn spotn lfot ohrwaaebtd oa tihlunbet e tdhm4oe0a m jvoaoorln dua eetmlMs. obaserpstw,h eetrehinec Snowbank Thickness ( 000...234 Obliquity = 25 component, CO , condenses during winter. In the 2 event that CO condenses, the surface energy balance 0.1 2 is calculated as discussed below. 0 We use the surface temperature to establish stability 0 5 106 1 107 1.5 107 2 107 2.5 107 3 107 3.5 107 4 107 Time (s) for CO2. As is common in these models, we calculate the condensation temperature of the atmosphere from Fig. 2. Sublimation proceeds more rapidly at the local pressure (700 Pa in all cases here). If the high obliquity, unless variations in temperature falls below the condensation atmospheric H O counteract this effect. 2 temperature, the mass of CO that condenses is 2 calculated by balancing the net radiation, the thermal emission, and the heat conducted upward into the cap from the underlying snow. The assumed albedo of CO2 is 0.4, which is Sublimation of the Seasonal Cap consistent with the observations of the cap, although 190 20 clearly the albedo of the cap is neither uniform nor 185 fixed. We use 0.4 because it represents an average of 180 15 te0Whm.9eh i.es osnbievsveiteryvr atothifeo t nhmse a rssesep aoosrfot enCdaO lf oC roO nt2h tech aeTp Eg irSso uadnsasdtua mi[s1 e>1d ] 0.t o,T whbeee Surface Temperature 111677505 10 2CO Stack 2 160 5 fix the wind speed (and hence the sensible and latent 155 heat fluxes), experienced by the snowbank at zero. 150 We also assume that the CO2 ice is opaque, so that no 20 20.5 21 21.5 220 solar insolation is absorbed by the snow once it is L s covered by CO . That is certainly a source of error Fig. 3. The effects of recession of the 2 when the seasonal cap is thin, particularly in the seasonal cap from the snowbank are spring. Indeed, in the case of cryptic regions in the captured. vicinity of the south pole, it appears that clear, transparent CO ice is sitting over a darkish surface. 2 It is not obvious however that clear CO ice could 2 form over snow. References: [1]. (Christensen, 2003). [2]. (Haberle, 1998). 3]Segura et al., 2002, [4]Hort and Weitz, 2002) [5]Jakosky and Carr, 1985,