Impact Craters as Hydrologic Refugia: Quantifying Subsurface Water-Ice Volumes in Martian Alluvial-Fan Hosting Basins using Terrestrial Analogues

The search and utilization of water ice is imperative for future Mars missions, underpinning life support, food production, propellant manufacture, and radiation shielding. Impact craters that host alluvial fans are key archives of Mars’ hydrologic transition and may also be prime subsurface water ice reservoirs. This study assesses their potential as accessible, high-volume water-ice “hotspots”.

Martian global surveys identify hundreds of craters with alluvial fans and deltas (314 craters, 890 fans, 206 fan-hosting craters; 1501 fan-shaped deposits in total), mostly Hesperian-Amazonian in age and concentrated from ~40ºN to 54ºS, especially in mid-latitudes and along the dichotomy boundary (Wilson et al., 2021; Morgan et al., 2022; Mondro et al., 2022). These fans record late-stage, often snowmelt fluvial activity under cold, generally arid climates (Wilson et al., 2021; Zhang et al., 2023). Many such basins coincide with regions of known or inferred excess ice, including mid-latitude mantles and basin-fill deposits (Wright et al., 2024; Dundas et al., 2022; Li et al., 2022).

Impact created structures strongly favor subsurface storage. Noachian-Early Hesperian craters and their basin fracture systems are proposed to dominate the ancient highland structural fabric and to have “groundwater enriched” and/or ice-rich zones that became trapped within a thickening cryolithosphere, wherever ground ice has been stable (Rodriguez et al., 2005). Hydrocode simulations show basin-forming impacts on Mars can generate several percent new porosity to depths of several kilometers and out to hundreds of kilometers from the impact, establishing enduring, fractures controlling fluid pathways and storage (Wiggins et al., 2022). Seismic studies further indicate a highly fractured, water-saturated upper and mid-crust capable of storing volumes equivalent to hundreds of meters to >1km of global equivalent layer (GEL), with porosity localized in fractures and crustal bedrock (Wright et al., 2024; Sun et al.,2025). Collectively, these results imply that impact fracture networks provide both the necessary porous host and permeable conduits to sequester large volumes of water and ice.

Building on this framework, the present work integrates a global catalog of 890 intracrater alluvial fans (Wilson et al., 2021) with impact-generated basin fractures and porosity concepts to evaluate fan-hosting craters as hydrologic “refugia” (Rodriguez et al., 2005; Wiggins et al., 2022; Cockell et al., 2024). Post-Noachian fans show strong equatorward flow biases in mid-latitudes, consistent with orographic cold-trap behavior where pole-facing rims accumulate snow and frost that subsequently melt and infiltrate beneath the surface. In this concept model, (i) Late Noachian surface water percolated downwards through the dense network of impact fractures before the upper crust cooled and froze (Sun et al., 2025), and (ii) later Hesperian-Amazonian alluvial fan development supplied additional meltwater that became trapped as ice within crater-floor and wall-fractures, analogous to groundwater enrichment of buried craters in the cryolithosphere (Rodriguez et al., 2005; Wilson et al., 2016). Large, ice-related subsidence and depressions in icy terrains (Jones et al., 2010; Sokołowska et al., 2024), ice-exposing impacts at 35ºN (Dundas et al., 2022), and orbital radar-inferred decameter-thick excess ice bodies (volumes of about 10⁴ km³) in mid-latitude basins (Wright et al., 2024) demonstrate that such impact basins can host substantial excess water ice.

To translate this qualitative picture into quantitative reservoir estimates (volumetric modeling), this study applies fracture-reservoir concepts from terrestrial analogs, where storage and longevity depend on fracture aperture distributions, connectivity, and the balance between high-permeability conduits and more diffuse damage zones. On Mars, simple and complex craters dominate total impact-fracture surface area and volume, vastly increasing potential space for fluid to percolate (Cockell et al., 2024; Wiggins et al., 2022). By combining alluvial-fan inventories, crater geometry, impact-induced porosity scaling, and available radar/topographic constraints, fracture volumetric modelling is used to estimate both bulk ice volumes and the potentially accessible fraction within fan-hosting basins.

This framework highlights intracrater alluvial-fan systems as double interest targets: they preserve paleoclimate records of Late Noachian through Amazonian aqueous activity, and due to impact-generated porosity and cold-trap accumulation, may overlie large, stratified reservoirs of large volumes of stored water ice sourced from both ancient oceans/lakes and younger alluvial fan percolated meltwater. Future radar sounding, seismology, and drilling focused on these craters could therefore simultaneously test models of impact-controlled hydrology, quantify storage volumes of ice, and identify accessible near-surface in-situ water resources for exploration.