Physically-based estimates of the age of ice in a martian debris-covered glacier

Introduction

Mars’ mid latitudes contain thousands of ‘viscous flow features’ (VFFs), akin to debris-covered glaciers on Earth [e.g. 1,2]. They are thought to have formed within the last several Myr to 100s Myr [3,4] during martian ‘ice ages’, driven by variations in Mars’ spin-axis obliquity [5,6]. Knowledge of the emplacement age of ice within VFFs is key to understanding the palaeoclimate histories likely contained within them.

Current VFF age estimates rely on the size-frequency distributions of impact craters across their surfaces [e.g. 3,4]. Such ages likely reflect the time since emplacement or last major modification of the surficial debris layer; they provide no direct information about the emplacement ages of the underlying ice, and implicitly assume a uniform age across the sampled area. However, ice flow physics causes spatial variations in ice flow, which will lead to the age of ice varying across VFF surfaces and with depth. Here, we develop a new physically-based approach to estimate the age of ice in VFFs, and apply this method to a small VFF in Mars’ southern mid-latitudes (Figure 1), the subject of our previous study [7].

Figure 1. VFF in Nereidum Montes (51.24°W, 42.53°S). (A) 6 m/pixel Context Camera (CTX) image P14_006572_1367_XN_43S051W showing the VFF (terminating at the white dashed line), and major arcuate VFF-surface structures cut through by a gully [8]. Inset: Mars Orbiter Laser Altimeter (MOLA) elevation map of Mars showing location of Nereidum Montes. (B) Oblique view of the VFF overlain by schematic map identifying key landscape features, adapted from [7]. Base image: orthorectified 25 cm/pixel HiRISE [10] image ESP_051036_1370, overlain on 1 m/pixel HiRISE DTM.

Methods

Our age-estimation method combines a 3-dimensional ice flow model [ISSM; 8] with particle tracking. We conduct a range of experiments with different assumed ice flow mechanisms with different exponents in the ice flow law (n = 2 or 3,  [9]), assumed VFF surface temperatures (T_s) ranging from 200 K to 230 K, and ice grain sizes from 0.5 mm to 5 mm. We take n = 3, T_s= 210K as the “standard” run. We track a dense network of 4860 “seed” particles across the VFF surface from emplacement, through transport within the VFF, to re-emergence onto the surface. Integrating the modelled velocity along each path allows calculation of the age of the surface at particle emergence points, and the age/depth relationship for any point within the VFF.

Results

Figure 2 shows the calculated VFF surface age for the standard run, with example particle paths and depths for 21 particles emerging along three of the major arcuate surface structures (Figure 1). There is a strong, rapid increase in surface age toward the VFF terminus; particle paths emerging nearer the terminus have come from furthest upstream (Figure 2B), and reach greater depths within the VFF. Vertical transects through the VFF (Figure 2C) show particle paths become increasingly tilted toward the ice surface toward the terminus (due to flow compression [8]); isochrons also become tilted upwards. The upstream area of the VFF shows younger ages; particles emplaced here move downwards into the VFF due to flow divergence, emerging at lower elevations (Figure 2B).

Figure 2. (A) Ice-surface age map for the standard run. Particle path depths shown as colour variations along 21 example paths. Red: model domain. White areas inside the domain with no age information: areas where no particle paths > 100 m long terminate, preventing age calculation. Hatched area: approximate portion of the domain where particle paths emergence points source from outside the domain northern edge. (B) Particle emergence points colour-coded by particle source elevation, for path lengths > 100m, and excluding paths emerging in the gully. Dashed lines: transects a–a’ and b–b’ (panel C). Background as Figure 1B. (C) Modelled Ice age/depth profiles for transects a–a’ and b–b’. Vertical flow paths emerging at the upper, middle, and lower surface structures shown for each transect. Ages shown at emergence points, and for hypothetical vertical transects 900 m from model domain westernmost edge. White areas as panel A.

Qualitatively, the spatial age patterns and particle paths are insensitive to the assumed ice flow mechanism, surface temperature and ice grain size. However, the latter two factors emerge as key controls on ice velocity and hence age. Mean ages of the 21 example tracks in the standard run (Figure 2) are (upper; middle; lower) 47.8 Myr; 82.3 Myr; and 178.4 Myr. These vary by ~100x with variable surface temperature, and by ~30x with grain size.

Conclusions

Our results show the age of near-surface ice varies significantly across the VFF, especially near the terminus. These patterns are driven by the 3-dimensional flow, particularly compression-driven upwards ice flow near the terminus bringing older, deeper ice to the surface. Age increases with depth throughout the VFF. Ice temperature and grain size, rather than the ice flow law, emerge as critical controls on ice velocity, and hence the specific calculated ice ages.

Our results have important implications for scientific sampling by future missions aiming to access ice, including robotic missions and eventual human-led missions which could extract climate records in ice cores. Our results highlight the importance of improving understanding of glacial processes on Mars ahead of ice access missions to identify potential landing sites and mission traverses with the highest science potential, and to devise ice sampling strategies.

References

[1]. Holt, J. W., et al. 2008. Science, 322, 1235–1238.

[2]. Plaut, J. J., et al. 2009. GRL, 36, 2008GL036379.

[3]. Baker, D. M. H., Carter, L. M. 2019. Icarus, 319, 264–280.

[4]. Hepburn, A. J., et al. 2020. JGR Planets, 125, e2019JE006102.

[5]. Laskar, J., et al. 2004. Icarus, 170, 343–364.

[6]. Madeleine, J.-B., et al. 2009. Icarus, 203, 390–405.

[7]. Butcher, F. E. G., et al. 2024. Icarus, 419, 115717.

[8]. Larour, E., et al. 2012. JGR: Earth Surf., 117, 2011JF002140.

[9]. Goldsby, D. L., Kohlstedt, D. L. 2001. JGR : Solid Earth, 106, 11017–11030.

[10]. McEwen, A. S et al. 2007. JGR: Planets, 112, 2005JE002605.