Lunar and Planetary Science XLVIII (2017) 2867.pdf CLIMATIC CONTROL OVER EXPLOSIVE VOLCANISM ON MARS. J. C. Andrews-Hanna1,2 and A. So- to1, 1Southwest Research Institute, 1050 Walnut St. Suite 300, Boulder, CO 80302, 2now at the Lunar and Planetary Laboratory, University of Arizona, Tucson, AZ 85721, [email protected]. Introduction and Background: On the Earth, One possible forcing mechanism is the periodic growth there is considerable evidence indicating both that vol- of thick ice deposits on the large equatorial volcanoes canic eruptions can drive climate change, and that cli- during period of high obliquity [22]. mate changes can modulate volcanism. Recent work Models and Results: As on Earth, possible mecha- on martian climate change has increasingly highlighted nisms for climatic modulation of volcanic activity on the important role that volcanic outgassing may have Mars include ice deposition/removal on the volcanoes, played in generating transient warmer and wetter cli- resulting in stress changes from glacial load- mates [1–4]. However, there has been little considera- ing/unloading and increases/decreases in the supply of tion of the possibility that climate changes may simi- meltwater to the magma chambers. larly have had an effect on volcanism. We first used the MarsWRF GCM [23] to examine Large volcanic eruptions on Earth are often corre- the formation of low-latitude ice deposits on the large lated with glaciations [5] or deglaciations [6, 7]. In volcanoes. This model includes surface albedo and some ash deposits, a regular ~23 kyr periodicity sug- thermal inertia effects, the full CO2 cycle, the effects of gests Milankovich forcing [8]. This influence of cli- dust, and a water cycle scheme that tracks vapor and mate over volcanism can be due to two key effects. ice in the atmosphere, ice on the surface, and vapor, ice First, the stress state of the lithosphere is altered by and adsorbate within the subsurface. We have run a loading or unloading associated with changes in sea low-resolution preliminary simulation, using a 5°×5° level or ice thickness. The decreasing stress in the edi- grid, initialized with 500 m water ice caps at both fice during unloading can trigger eruptions [9, 10]. poles, an obliquity of 45°, and current eccentricity and argument of perihelion. After five years, the climate is Second, climate-induced increases in water supply to redistributing water ice from the polar regions into the the magma chamber and conduits can trigger explosive low latitudes. Water ice preferentially deposits on the volcanic activity. A number of volcanic dome collapse western flanks and summits of the volcanoes in the and pyroclastic eruptions have been linked to rainfall Tharsis region (Fig. 1). The distribution of ice is broad- and/or glacial melt [11–13]. Here we examine the pos- ly similar to previous work [22], but favors deposition sible role that climate changes may have played in on the more northern Olympus and Ascraeus Montes, modulating explosive volcanism on Mars. rather than the more southern Arsia and Pavonis Mon- There is abundant evidence for both effusive and tes. While the maximum ice deposition rate occurs on explosive volcanic activity throughout martian history the flanks, significant deposition occurs in the calderas [14]. The Medussae Fossae Formation (MFF) is com- as well. monly interpreted as a thick accumulation of volcanic ash [15], likely sourced from a combination of Apolli- naris Mons and the Tharsis Montes [16, 17]. Pyroclas- tic deposits are also found on Olympus Mons and the Tharsis Montes [18]. In comparison to Earth, basaltic plinian eruptions may be more common on Mars due to the lower gravity and atmospheric pressure [14]. Possible observational evidence for climatic control over volcanic activity on Mars is found in sedimentary layering exhibiting a highly regular periodicity in layer thickness [19]. In at least one example, the layers are bundled into repeating packets of 10 [20], correlating with the modulation of the 120 kyr obliquity cycle on a Figure 1. Ice accumulation in the Tharsis Montes region for an obliquity of 45° and present-day eccentricity and argu- 1.2 Myr timescale [21]. Some examples of strongly ment of perihelion. The results are smoothed from the 5° periodic layering are found in unaltered ash deposits, resolution of the preliminary GCM, and superimposed over including the MFF and the upper formation of Mount MOLA shaded relief for context. Sharp in Gale Crater [19]. The observation of highly regular periodicity in thick stratigraphic sections with- One mechanism for triggering explosive volcanic in martian ash deposits supports a periodic forcing activity is by generating a flux of water to the magma mechanism, as can be provided by the climate system. chamber or conduits before or during an eruption. Lunar and Planetary Science XLVIII (2017) 2867.pdf Since liquid water has been scarce on Mars in compar- in modulating volcanic eruptions. Geological evidence ison with Earth, the effect of an increase in water sup- from layered ash deposits on Mars suggest a similar ply may be more pronounced on Mars. The heat flow link between periodic orbitally induced climate chanc- and melt generation within Olympus Mons was mod- es and explosive volcanic eruptions. GCMs confirm eled using a finite difference model in an axisymmetric that thick ice deposits should form on the equatorial spherical cap geometry. The model surface topography volcanoes during periods of high obliquity. Models of was taken from a MOLA profile of Olympus Mons. the thermal and stress responses to this ice loading and The background heat flux includes both bottom heating unloading show that large volumes of melt water can and internal radiogenic heating. The magma chamber be generated at the base of caldera ice deposits overly- was represented as a constant-temperature boundary ing active magma chambers, and that the tensile stress- condition, assuming a magma temperature of 1270 K. es during glacial unloading are sufficient to trigger The dynamic thermal response to periodic ice deposi- volcanic eruptions. Interestingly, the lag in warming of tion was modeled by adding and removing cells to the top boundary representing the ice deposits as a simple the base of the ice sheet resulting from the thermal cosine-function of time. with a maximum deposition diffusion timescale results in peak melt rates being rate of 50 mm/yr. For a magma chamber top depth of sustained up through the time of peak ice thinning by 10 km and radius of 20 km, the model predicts sub- sublimation loss, making the maximum rate of melt stantial melt generation, with a peak melt thickness of production contemporaneous with the maximum ten- ~710 m and a peak melt flux of ~1 cm/yr (Fig. 2). sile stresses from unloading. These results suggest that orbitally induced climate changes on Mars may have a played an important role in modulating the timing and 50 b 2.0 ice+water water style of volcanic eruptions, encouraging explosive vol- 1.5 accumulation canic activity during periods of high obliquity and con- 0 ablation 1.0 0.5 tributing to the periodicity of layering within thick P-E (mm/yr) −50 thickness (km) 0 accumulations of ash. 0 20 40 60 80 100 120 0 20 40 60 80 100 120 time (kyr) time (kyr) c Temperature (K) d dT/dz (K/km) References. [1] Phillips R. J. et al. (2001) Science. 291 100 2587–2591. [2] Halevy I. and Head J. W. (2014) Nat. 1400 80 Geosci. 7 865–868. [3] Mischna M. A. et al. (2013) 1200 JGR 118 560–576, doi: 10.1002/jgre.20054. [4] Urata 60 R. A. and Toon O. B. (2013) Icarus 226 229–250. [5] 800 40 Rampino M. R. and Self S. (1993) Quarternary Res. 600 40 269–280. [6] Nowell D. A. G. et al. (2006) J. Quat. 20 Sci. 21 645–675. [7] Glazner A. F. et al. (1999) GRL 400 26 1759. [8] Paterne M. et al. (1990) EPSL 98 166– 0 Figure 2. Assumed mean annual net accumulation/ablation 174. [9] Allan A. S. R. et al. (2008) Quat. Sci. Rev. 27 (a), and resulting thicknesses of ice and melt water (b) from 2341–2360. [10] Jellinek A. M. et al. (2004) JGR 109 the thermal model. Snapshots of the temperature and temper- B09206, doi:10.1029/2004JB002978. [11] Matthews ature gradient are shown in c and d (ice and magma chamber A. J. (2002) GRL 29 1644, doi:10.1029/ are outlined in white and black; vertical exaggeration is ~5x). 2002GL014863. [12] Capra L. (2006) J. Volcanol. Geotherm. Res. 155 329–333. [13] Deeming K. R. et We next examined the effect of unloading of the al. (2010) Philos. Trans. A. Math. Phys. Eng. Sci. 368 edifice during thinning of the ice deposits. We used the 2559–2577. [14] Wilson L. and Head J. W. (1994) TEKTON finite-element model to represent the re- Rev. Geeophys. 32 221–263. [15] Carter L. M. et al. moval of an ice deposit following a quarter-cosine (2009) Icarus 199 295–302. [16] Kerber L. et al. function from the center of the edifice to the top of the (2012) Icarus 219 358–381. [17] Hynek B. M. (2003) basal scarp. The preliminary results show a radial ex- JGR 108 5111, doi:10.1029/2003JE002062. [18] tensional stress at the surface of 5.6 MPa. Stress Bleacher J. E. et al. (2007) JGR 112 E04003, changes within the edifice due to ice unloading signifi- doi:10.1029/ 2006JE002826. [19] Lewis, K. W., Ahar- cantly exceed the critical stress required to trigger onson, O., (2014), JGR 119, 1432–1457. [20] Lewis K. eruptive activity [10]. Interestingly, the thermal lag W.