The origin of sinuous ridges in Argyre Planitia: Insights from terrestrial analogues and implications for its hydrology

The Argyre basin, located on the southern highlands, spans over 1500 km wide and is one of the largest impact basins on Mars. Three large valleys drain into Argyre from the southern circumpolar region, whereas the basin is breached at the north and perhaps connected to the Ladon-Morava-Ares outflow system draining into Chryse Planitia. Given its volume and valley connections, this basin played a key role in the global hydrology of early Mars (e.g., Clifford and Parker, 2001; Phillips et al., 2001). Several morphological features indicate that the basin may have contained a lake and/or ice sheet in the past, which could have been fed with meltwater from an ancient south polar ice sheet through three main inlet valleys (e.g., Hiesinger and Head, 2002; Ghatan and Head, 2004). A suite of sinuous ridges in the southern Argyre basin (ASRs)—commonly regarded as eskers—has been interpreted as evidence that an ice sheet once occupied the basin (e.g., Banks et al., 2009; Bernhardt et al., 2013). However, new morphological observations focusing on the context and stratigraphy raise discrepancies with the conventional esker interpretation, leading us to revisit their origin.

To assess the origin of these ridges, we first have undertaken detailed morphometric measurements of ASRs and compared them to other glacial and fluvial ridge features on Earth and Mars. We mapped cross-sectional profiles at 1 km intervals along crestlines of ASRs using Context Camera (CTX) images and digital elevation models (DEM). We fitted power-law relationships to those ridge geometries, including width (W) and cross-sectional area (A_CS). A power law is expected here because differing scaling relationships exist between channel width and bank-full discharge in fluvial (e.g., Parker et al., 2007) and subglacial systems (e.g., Ng, 1998, Hewitt and Cretys, 2019). We use A_CS as a proxy for bank-full discharge (e.g., Ruso et al., 2024). Next, we carefully considered the stratigraphy of the ridges as well as the context where they are found, characterized by laterally extensive layered terrains. We measured dip and azimuths of individual layers observed within the ridge stratigraphy and compared them to those in the surrounding terrains. Then, we worked to understand the stratigraphic relationships between ASRs and surrounding layered terrains. Last, we performed crater counts using the buffer crater counting technique on the ASR to determine their ages (e.g., Kneissl et al., 2011).

Two distinct populations of ASRs can be distinguished stratigraphically: NE-oriented ridges in the eastern population (upper) and NW-oriented ridges in the western population (lower). The power-law relationship for ASRs geometry, shown in Figure 1, shows a strong correlation between log-width and log-cross-sectional area, which is consistent with martian inverted fluvial channels in Aeolis Dorsa region, though the power-law exponents for ASRs are smaller than, yet still comparable to, the subglacial range. The measurement of layering structures revealed that the layers in ASRs are extremely horizontal (~0.07–<0.3°; Figure 2a). These values of dip and azimuth are consistently observed both in ASRs and the surrounding terrains, suggesting a shared sedimentological history, consistent with inverted fluvial channels where the former protective cap rocks have been eroded—manifesting as ridges composed of underlying shale bedrock extending from adjacent lacustrine or floodplain units (e.g., in Utah - Figure 2b). The lateral continuity of layers between ridges and surrounding terrain is entirely inconsistent with the hypothesis of eskers previously proposed (Banks et al., 2009). Whereas eskers may be layered, they are confined structures, preventing them to deposit simultaneously with the surrounding terrain, as subglacial conduits tend to draw water from the surrounding water-distributed regions because the water pressure in a conduit is less than in the adjacent bed. Thus, we conclude that the origin of ASRs is inverted fluvial channels, and the surrounding layered terrains are lacustrine deposits. We have dated ASRs to ~3.6 Ga as exhumed age. Given that the Argyre impact age is ~3.9 Ga, the basin may have hosted a lake at approximately the same time as ASRs (~3.8–3.6 Ga).

Furthermore, we also identified similarly layered terrains in other regions of the Agyre basin, justifying an extrapolation of the horizontal layers to the entire basin. The elevation of the ASRs and their layers ranges from approximately −2633 to −2734 m, and the layers extend to even lower elevations. This elevation indicates the minimum lake level of the putative paleo-Argyre lake, which remains below the hypothesized breach point at the northern end for an Argyre–Uzboi Vallis flood. The two distinct populations of ASRs could provide possible evidence for multi-fluvial periods driven by meltwater from an ancient southern polar ice sheet (e.g., Ghatan and Head, 2004), with the Argyre basin serving as an impoundment for the meltwater (lake volume: ~1.2 × 10⁵ km³).

Figure 1: Comparative geometry for ridge features. (a) ASR compared with martian inverted fluvial channels in Aeolis Dorsa region. (b) ASR compared with martian eskers in Dorsa Argentea Formation.

Figure 2: 3D view of layering structures. Black arrows indicate ridge landforms. (a) ASRs and surrounding layered terrains. CTX_669660_1229 and CTX_067573_1245. Colored lines indicate layers which appear both ASRs and surrounding layered terrains. (b) analogous terrestrial inverted fluvial channels where the cap rocks have been eroded, Utah, US (38.393565°N, 110.788020°W).