History and anatomy of subsurface ice on Mars Norbert Schorghofer a,⇑ , Francois Forget b a Institute for Astronomy, 2680 Woodlawn Drive, University of Hawaii, Honolulu, HI 96822, United States b Laboratoire de Météorologie Dynamique, Université Paris 6, Paris, France article info Article history: Received 9 November 2011 Revised 1 July 2012 Accepted 4 July 2012 Available online 16 July 2012 Keywords: Mars, Climate Mars, Surface Mars, Polar Geology abstract Ice buried beneath a thin layer of soil has been revealed by neutron spectroscopy and explored by the Phoenix Mars Lander. It has also been exposed by recent impacts. This subsurface ice is thought to lose and gain volume in response to orbital variations (Milankovitch cycles). We use a powerful numerical model to follow the growth and retreat of near-surface ice as a result of regolith–atmosphere exchange continuously over millions of years. If a thick layer of almost pure ice has been deposited recently, it has not yet reached equilibrium with the atmospheric water vapor and may still remain as far equatorward as 43°N, where ice has been revealed by recent impacts. A potentially observable consequence is present- day humidity output from the still retreating ice. We also demonstrate that in a sublimation environ- ment, subsurface pore ice can accumulate in two ways. The first mode, widely known, is the progressive filling of pores by ice over a range of depths. The second mode occurs on top of an already impermeable ice layer; subsequent ice accumulates in the form of pasted on horizontal layers such that beneath the ice table, the pores are completely full with ice. Most or all of the pore ice on Mars today may be of the sec- ond type. At the Phoenix landing site, where such a layer is also expected to exist above an underlying ice sheet, it may be extremely thin, due to exceptionally small variations in ice stability over time. Ó 2012 Elsevier Inc. All rights reserved. 1. Introduction The distribution of subsurface ice on Mars measured with neu- tron and gamma-ray spectroscopy (Boynton et al., 2002; Feldman et al., 2002; Mitrofanov et al., 2002) is consistent with equilibrium model calculations based on exchange of water vapor between the atmosphere and the subsurface (Mellon et al., 2004; Schorghofer and Aharonson, 2005; Diez et al., 2008). On a zonal average, sub- surface ice is expected poleward of about ±55° latitude on both hemispheres. Recent progress in subsurface ice modeling (Schorghofer, 2007a, 2010) enables us to not only determine the equilibrium dis- tribution but follow the evolution of the ice in response to climate change. Equilibrium and non-equilibrium models both predict that the size of the two hemispheric ice layers has changed over the past few million years (Mellon and Jakosky, 1995; Schorghofer, 2007a) in response to changes in the orbit and in axis orientation (Ward, 1992; Laskar et al., 2004). Deposition and desiccation of an ice-rich mantle is also suggested by geomorphic evidence (e.g., Mustard et al., 2001; Kreslavsky and Head, 2002; Head et al., 2003; Kostama et al., 2006). The subsurface ice can be emplaced in two ways. (1) Snowfall or direct deposition during a past climate period, when the obliquity (axis tilt) was different, may have led to the formation of a peren- nial snow cover that subsequently densified. During retreat of this ice, any dust in the ice would remain as sublimation lag, leading to self-burial. (2) The second emplacement mechanism, uncommon on Earth, is subsurface ice directly deposited from the vapor phase. It fills the available voids between soil grains and is thus called ‘‘pore ice’’ or interstitial ice. This vapor deposition process has been predicted by Mellon and Jakosky (1993) and it has been repro- duced in the laboratory by Hudson et al. (2009). Here, a dynamic (i.e., non-equilibrium) model is used to follow the evolution of near-surface ice with time. We assume specific ini- tial conditions (such as an ice sheet of given thickness and extent) and surface humidity history (based on GCM results) and follow how the near-surface ice involves as a result of atmosphere– subsurface exchange of water vapor. Pore ice is allowed to form anywhere that conditions allow. These model calculations will pro- vide quantitatively plausible scenarios that connect initial condi- tions with the present-day near-surface ice distribution. We arrive at three new results: 1. When ice-cemented soil forms from atmospherically derived water vapor, it can do so not only by gradual growth over a range of depths below an ice table, but also by vertical upward movement of the ice table, without intermediate stages of pore- filling (Section 4). 2. At the Phoenix landing site, model calculations with an initial ice sheet predict three layers: dry soil, ice-cemented soil, and 0019-1035/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.icarus.2012.07.003 ⇑ Corresponding author. E-mail address: norbert@hawaii.edu (N. Schorghofer). Icarus 220 (2012) 1112–1120 Contents lists available at SciVerse ScienceDirect Icarus journal homepage: www.elsevier.com/locate/icarus