THE FORMATION OF MARTIAN DUNE GULLIES BY DRY ICE ... · at Noachis and Aonia Terrae, Mars. In Planetary Dunes Workshop, pages 27–28, 2008. [3]D. Reiss, R. Jaumann, A. Kereszturi,

THE FORMATION OF MARTIAN DUNE GULLIES BY DRY ICE — EXPERIMENTS AND MODELLING.Jim McElwaine1,2, Serina Diniega3, Candice Hansen1, Mary Bourke1,4 and Joanne Nield5. 1Planetary Science Insti-tute, Tucson, USA, (jmcelwaine@psi.edu), 2Durham University, UK, 3JPL, Pasadena, USA, 4Trinity College Dublin,Ireland, 5University of Southampton, UK.

Figure 1: Zig-Zag gullies in a Dune Field in Kaisercrater, HiRISE PSP 010749 1325

Introduction Long, narrow grooves found on theslopes of martian sand dunes were first reported by Man-gold et al. [1] and are most likely the result of largeblocks of dry ice (figure. 1. Imaging by the Mars OrbiterCamera (MOC) on the Mars Global Surveyor and theHigh Resolution Imaging Science Experiment (HiRISE)on the Mars Reconnaissance Orbiter has demonstratedthat these linear gullies are found within many dunefields and on sandy crater walls within the mid-latitudeson pole-facing slopes [2, 3]. These slopes typically rangefrom 7 to 12◦ (well below the angle at which a dry granu-lar material is expected to flow [4, 5, 3]), but the gully al-coves and grooves appear to originate within the steeperupper slope [can be > 25◦; 4]. Over the past six Marsyears, HiRISE images show that existing grooves haveelongated and new grooves have formed at the start ofeach spring, demonstrating that these features are activein the present-day Martian climate.

The dry ice block hypothesis is consistent with theobserved morphology, location, and current activity: thatblocks of carbon dioxide ice break from over-steepenedcornices as sublimation processes destabilize the surfacein the spring, and these blocks move downslope, carv-ing out leveed grooves of relatively uniform width andforming terminal pits.

Fieldwork: To test this hypothesis, we have per-formed experiments at two dune fields, Grand Falls inArizona and Coral Pink in Utah. Dry ice blocks werereleased and on a variety of slopes and their positions

Figure 2: Setup on a falling dune at Grand Falls. Twohigh resolution video cameras, two calibration targetsand a Terrestrial Laser Scanner are visible.

Figure 3: Ripples are gradually smoothed out by the re-peated passing of the CO2 blocks.

tracked in three dimensions using video cameras. In ad-dition Terrestrial laser scanning was used to create dig-ital terrain models and to map the changing morphol-ogy of the surface (figure 2). The data is combined toproduce trajectory data for each block measured as arclength down the thalweg as a function of time. The re-sults show that steady movement is possible on slopes ofas little as five degrees.

Modelling: A model for the levitation of CO2 blockswas developed in [6]. The flow of CO2 gas within thesand bed is given by Darcy’s law. The heat flux is calcu-

3035.pdfFifth Intl Planetary Dunes Workshop 2017 (LPI Contrib. No. 1961)

lated by solving a heat diffusion equation which have theresult that, with Martian parameters, a block will levitatewhen

t < t∗H

R=H

R

(πµh

4gkeρsρ

)2

= 2.9H

Rs,

where ρs is the bulk density of sand, c is the heat ca-pacity and κs is the thermal conductivity of sand. Thuseven blocks of high aspect ratio (H/R) can levitate fora few seconds. The extended model had the same ba-sic equations for heat flow and CO2 flow, but when theblock is levitating the heat flux will be much lower. Con-duction now takes places through a combination of solidconduction and gas phase conduction with a linear tran-sition depending on the CO2 pressure relative to the solidpressure. The temperature boundary condition under theblock then becomes

h =Tz=0 − Ts

r

[κs +

pz=0

Hρsg

Tz=0 − Tsr

(κg − κs)

],

where κg is the thermal conductivity of CO2. This heatflux under the block now depends on the position, sincethe pressure is spatially varying and a numerical solutionis required. As the block moves it also encounters freshsand which is hotter enhancing the mobility.

Finally there is an equation for the motion of theblock. It is driven down the slope by gravitymg sin θ andresisted by a frictional force proportional to the blocksweight minus the supporting CO2 pressure. Finally thereis a form drag due to ploughing like interactions with sur-face roughness. Taking s to be the coordinate down thethalweg (steepest descent slope) the equation of motionis

ms = mg sin θ − µ(mg cos θ − F )− 12ρsRrs

2,

where F is the CO2 pressure integrated under the block,C is a drag coefficient width and r a roughness lengthcharacterising the surface. The surface roughness r de-pends on the initial condition of the sand surface and thendevelops as blocks are released and remodel the surface.If there are initially ripples the surface roughness is highand reduces over time (figure 3). If the surface is initiallysmooth the blocks induce surface undulations similar towashboard road or moguls (figure 4).

Acknowledgements: Funded by NASA MFRPNNH13ZDA001N. Sincere thanks to the Navajo Nation,the State of Utah and Tim Titus.

Figure 4: The initially smooth surface develops a wah-board pattern after multiple passings.

References:[1] N Mangold, F Forget, and F. C. J.-P. Peulvast. Narrow Gul-

lies Over High Sand Dunes On Mars: Evidence For RecentLiquid Flows. In EGS General Assembly Conference Ab-stracts, volume 27 of EGS General Assembly ConferenceAbstracts, page 3080, 2002.

[2] G Di Achille, S Silvestro, and G.˜G. Ori. Defrosting Pro-cesses on Dark Dunes: New Insights from HiRISE Imagesat Noachis and Aonia Terrae, Mars. In Planetary DunesWorkshop, pages 27–28, 2008.

[3] D. Reiss, R. Jaumann, A. Kereszturi, A. Sik, andG. Neukum. Gullies and Avalanche Scars on Martian DarkDunes. In Lunar and Planetary Institute Science Confer-ence Abstracts, volume 38 of Lunar and Planetary Inst.Technical Report, page 1993, March 2007.

[4] G Jouannic, J Gargani, F Costard, G.˜G. Ori, C Marmo,F Schmidt, and A Lucas. Morphological and mechani-cal characterization of gullies in a periglacial environment:The case of the Russell crater dune (Mars). Planet. Space.Sci., 71:38–54, October 2012.

[5] N. Mangold. Debris flows over sand dunes on Mars: Ev-idence for liquid water. Journal of Geophysical Research,108(E4):5027, April 2003.

[6] S Diniega, C.J. Hansen, J.N. McElwaine, C.H. Hugen-holtz, C.M. Dundas, A.S. McEwen, and M.C. Bourke. Anew dry hypothesis for the formation of martian linear gul-lies. Icarus, 225(1):526–537, July 2013.

3035.pdfFifth Intl Planetary Dunes Workshop 2017 (LPI Contrib. No. 1961)