Martian gullies, characterized by alcoves, channels, and aprons, have sparked debate regarding their formation mechanisms. The prevailing hypotheses involve liquid water, brines, or sublimating CO₂ ice, or a combination. While gully morphology closely resembles aqueous debris flows on Earth, present-day activity is inconsistent with the scarcity of liquid water on Mars. Evidence of ongoing activity includes new lobate deposits, channel incision, boulder movement, and channel bank collapse, indicating active landscape shaping rather than mere modification of pre-existing landforms. These flows exhibit fluidization, and the timing of activity correlates with CO₂ frost presence, strongly suggesting the involvement of sublimating CO₂. Existing CO₂-related hypotheses include fluidization of sediment-CO₂ mixtures by gas flux and sediment transport by pressurized flows beneath sublimating CO₂ ice slabs. However, the slab ice hypothesis is limited to higher latitudes. Two key challenges remain: the lack of in-situ observations of gully activity and an incomplete understanding of climatic and topographic conditions that facilitate CO₂-driven flows. This study addresses these challenges by presenting a new framework validated with the first physical tests of CO₂ sublimation-driven granular flows under Martian atmospheric conditions and a climate model to assess the spatial and temporal distribution of these flows.

Numerous studies have explored the formation of Martian gullies. Early research focused on liquid water and brines as the primary drivers (e.g., Malin & Edgett, 2000; Costard et al., 2002; Knauth & Burt, 2002). However, the low availability of liquid water under current Martian conditions has prompted alternative explanations emphasizing the role of sublimating CO₂ frost (e.g., Christensen, 2003; Conway et al., 2011). Several studies presented observational evidence of present-day gully activity linked to seasonal CO₂ frost (e.g., Diniega et al., 2010; Dundas et al., 2010, 2012, 2015; Raack et al., 2015; de Haas et al., 2019; Khuller et al., 2021; Dundas et al., 2022; Pasquon et al., 2023; Sinha & Ray, 2023). These studies highlighted the morphological similarities between Martian gullies and terrestrial debris flows, but also emphasized the challenges in reconciling the observed activity with the limited availability of liquid water. Hypotheses regarding CO₂-driven mechanisms involved fluidization by sublimating CO₂ or subsurface pressurized flows beneath CO₂ ice slabs (e.g., Pilorget & Forget, 2016). However, a comprehensive framework integrating experimental and modeling approaches to test the feasibility and conditions under which CO₂ sublimation drives present-day gully activity was lacking.

This research employed a two-pronged approach: laboratory flume experiments and 1D climate modeling. Flume experiments were conducted in a Mars chamber at the Open University, simulating CO₂-driven granular flows under Martian atmospheric pressure (800 Pa). The flume setup, inspired by terrestrial debris flow studies, consisted of a sediment reservoir, a chute with a 30° slope, and an outflow plain. Heating pads controlled the chute floor temperature. Two laser distance sensors measured flow velocity. Four key experiments, each with duplicates, tested two proposed mechanisms: (1) a mixture of sand and CO₂ ice flowing over a heated surface; and (2) dry sand flowing over a CO₂ ice layer. Reference experiments excluded CO₂ ice (under Martian pressure) and used CO₂ ice under Earth's atmospheric pressure. The sediment used was a moderately sorted sub-rounded fluvial sand. A 1D atmospheric model, a simplified version of the Mars Planetary Climate Model (PCM), predicted the spatial and temporal distribution of CO₂ frost and temperature differences on slopes, simulating conditions necessary for the two mechanisms. The model accounted for solar irradiance, CO₂ condensation/sublimation, and soil thermal inertia (TI). Three TI values (750, 1000, and 1250 J m⁻² s⁻¹ K⁻¹) were used to represent different soil compositions. The model was tuned to match observed CO₂ frost occurrence on steep slopes, necessitating higher TI values than typically used in Martian climate modeling. The model results were then compared to the observed locations and timing of present-day gully activity.

The flume experiments provided direct evidence that CO₂ sublimation can fluidize and sustain granular flows under Martian conditions. Both tested mechanisms resulted in strongly fluidized flows with high velocities (>2 m s⁻¹), multiple surges, increased flow depth, and elongated depositional lobes. Small levees in the deposits indicated grain segregation and complex velocity profiles. Fluidization was most pronounced in mechanism 2 (sand over CO₂ ice). The reference experiments (no CO₂ under Martian pressure and CO₂ under Earth pressure) showed no fluidization. Dimensional scaling analysis showed that less CO₂ volume flux is needed for fluidization under Martian gravity compared to Earth. The 1D climate model predicted that the conditions required for both mechanisms (coexistence of frosted and defrosted slopes with sufficient temperature difference) can occur on Mars at various latitudes and times of year. The model's predictions for the locations and timing of gully activity generally aligned with observed data, particularly for mechanism 1 and south-facing slopes in mechanism 2. The model also predicted the inactivity of some gully populations (e.g., northern hemisphere gullies at lower latitudes). The observed deposit morphologies in the experiments (levees and lobes) resemble features in Martian gully deposits, indicating similarities in flow dynamics. The study concludes that CO₂ sublimation can effectively mobilize and transport sediment, even over low gradients, and that the energy required for sublimation can be obtained from warmer adjacent slopes, eliminating the need for liquid water to explain present-day Martian gully activity.

The findings of this study challenge the long-held assumption that the presence of gullies on Mars is definitive proof of past or present liquid water activity. The combined results of the laboratory experiments and climate modeling strongly support the hypothesis that CO₂ sublimation can be a primary driver of present-day gully activity. The close correspondence between the modeled climatic conditions and the observed distribution of active gullies provides compelling evidence for the validity of this mechanism. The lack of fluidization in the control experiments, where CO₂ was absent or atmospheric pressure was higher, underlines the crucial role of CO₂ sublimation under Martian conditions. The observed similarities between the experimental deposit morphologies and Martian gully features further strengthen the proposed mechanism. This research significantly alters our understanding of Martian geomorphology and has implications for the search for evidence of past or present water on Mars, as it suggests that less water may be needed to explain observed features than previously thought.

This research demonstrates, through a combination of laboratory experiments and climate modeling, that CO₂ sublimation is a viable, and potentially dominant, mechanism for present-day Martian gully activity. The findings challenge the long-standing assumption that gully morphology necessarily indicates the presence of liquid water. The model accurately predicts the location and timing of active gullies based on CO₂ sublimation, while also explaining inactive gullies in specific regions. Future work could focus on more sophisticated modeling incorporating factors like subsurface water ice and infrared radiation, in-situ observations to confirm the processes, and extending the framework to other planetary bodies with low atmospheric pressure to re-evaluate the interpretation of gully-like features.

The study's limitations include the use of a 1D climate model, which simplifies some aspects of the Martian climate. The model does not fully account for the effects of subsurface water ice or incident infrared flux on slope temperatures. Furthermore, the experiments were conducted at a relatively small scale, and the extrapolation to natural Martian conditions requires careful consideration. While the study provides strong evidence for CO₂ sublimation as a significant driver of gully formation, in-situ measurements of gully activity and detailed sedimentological analysis would provide definitive confirmation.