Evidence for a transition from a long-lived "cold and icy" climate to a short-lived "warm and wet" climate in the Late Noachian

Geologic features on the surface of Mars are suggestive of a "warm and wet" Late Noachian (LN) climate, characterized by above-freezing surface temperatures and abundant fluvial and lacustrine activity from rainfall and runoff. The inability of climate models to produce continuous and long-lived "warm and wet" conditions, paired with the lack of these features on older terrains, suggests that this warm climate may have been confined to the LN, not continuous throughout the Noachian.

Recent studies have shown that short-lived "warm and wet" conditions could be achieved through greenhouse warming from an atmosphere enriched in reducing gases, such as H₂or CH₄. However, in model-simulated climates with global mean annual temperature ~273 K, a reasonable surface water inventory, and present-day topography, significant portions of the Tharsis rise and south polar region are below freezing year-round and, as a result, are surface cold traps. Rainfall is minimal, precipitation is dominated by snowfall, and fluvial activity is dominated by snowmelt and runoff, not rainfall.

Thus, we explore the influence of snowmelt for driving fluvial activity and forming the valley networks (VN). In this study, we use the 3D LMD GCM to assess whether (1) predicted snowmelt distributions in a "warm and wet" climate are consistent with the distribution of VN and, thus, VN can be explained by a continuous and long-lived "warm and wet" Noachian climate, or (2) predicted snowmelt distributions in a "warm and wet" climate are not consistent with the distribution of VN. In this alternative case, could significant melting across the highlands have occurred during the transition from a "cold and icy" to a "warm and wet" climate? Does the distribution of VN support evidence for a long-lived "cold and icy" Noachian climate with a period of transient warming in the Late Noachian?

Our statistical analysis of GCM results is consistent with the second scenario; VN distributions are consistent with the distribution of predicted ice melting during a transition from a "cold and icy" to a "warm and wet" climate. Important implications for the Noachian climate are that the "warm and wet" period may have been short-lived in the in the Late Noachian and fluvial activity may have been dominated by snowmelt instead of rainfall.