Water and lava both seem viable for the formation of one of Mars' densest and largest channel networks

The Axius Valles on the Malea Planum region’s (MPR) northern flank down into the Hellas basin are one of the most extensive and densest channel networks on Mars [1,2]. While previous studies tentatively interpreted the area as pyroclastic deposits dissected by sapped water/lahar flows [3-8] we considered their viability versus low-viscosity lava flows. 

Physiography
The Axius Valles and adjacent channels to the west consist of ~22,550 km of mostly parallel sinuous valleys dissecting a plain (drainage density of 0.09 km^-1) of gentle but relatively uniform northnortheast tilt, i.e., long-wavelength dip, at ~0.6 to 0.9° (~1 to 1.6%) towards the Hellas basin floor. The channels are up to ~20 km wide and ~100 m deep, although most are narrower and shallower than ~5 km and ~50 m, respectively. The majority of the valleys originates around the rim of Amphitrites Patera between elevations of ~1,200 and ~1,600 m. Smaller subsets originate at or below the rim of Peneus Patera between elevations of ~0 m and ~600 m, or are traceable further south into the wrinkle-ridged plains of the MPR. The longest continuously traceable valley of the Axius Valles is ~325 km long and follows the topographic gradient from ~600 m above the datum down to ~ -4,800 m. The valleys’ sinuosity is relatively low, ranging from ~1 up to ~1.15, and anabranching is very common. In several locations, sinuous valleys are levéed, i.e., bound by ridges that can be up to ~100 m high. 

Discussion
Based on their morphology and location, the Axius Valles have been tentatively interpreted as the result of sapped water or lahar flows that carved into friable pyroclastic deposits [3-7]. However, diagnostic features such as short, digitate levée-overspill deposits, bulged, lobate flow fronts (both typical for high viscosity flows, i.e., most lavas or mud/sludge), and associated pit-chains (typical for lava tubes, i.e., lava flows) are absent but might have been covered by 10s of meters thick dust-ice mantling [12] or eroded by intense deflation [e.g., 13]. In any case, the fact that the channels extend over 100s of kilometers on a slope of <1° seems to favor low viscosity density currents. Water or sludge flows stand to reason especially as ice accumulation models for an ancient martian 1 bar atmosphere predict a several 10s of meters thick ice sheet, i.e. potential melt water source, to form on the highest points of Amphitrites Patera [14]. Nevertheless, due to the geographic association with this patera – likely one of the largest calderas on Mars [e.g., 11,15] – the plausibility of very low-viscosity lavas such as komatiite and tholeiitic basalt [16-18] as alternatives to water should be ascertained. Mantle-derived low-viscosity magmas such as komatiite or tholeiitic basalt [e.g., 19-21] are indicated by the broad and gently sloped shields of Amphitrites and Peneus Paterae (11,15,21] and also an expected product of MPR volcanism, which was likely caused by deep ring-fractures and mantle upwelling related to the Hellas basin-forming event [11,15]. Furthermore, models indicate that komatiite and tholeiitic basalt flows on very shallow slopes should be able to travel up to ~325 km and form ~100 m deep channels if flow durations and 2-dimensional discharge rates are at least several months and ~150 m² s^-1, respectively [22,23]. In the channels close to the patera summits, whose average width is ~3 km, this would result in a discharge rate of 450,000 m³ s^-1, which is within the spectrum deduced for other large terrestrial, lunar, and martian flows [24,25]. Given the sizes of Amphitrites and Peneus Paterae as potential source areas, as well as the volume of potentially basaltic material filling the Hellas basin (~10⁶ km³ [13]), such discharge rates might be feasible, especially as itwould be a peak value and not constant over the course of a months-long eruption. Lastly, as is the case for overlapping and interacting lava channels on Earth, e.g., on the flanks of Etna or Teide, such networks form sequentially and not all at once, thereby suggesting a volcanic formation of the Axius Valles would have included multiple eruptions, too.

Preliminary Conclusions
The primary parameters of the Axius Valles, i.e., their sinuosity, size, anabranching, levées, and drainage density are not diagnostic and could be explained by multiple types of density currents. The channels’ length over a gentle slope implies low-viscosity liquids, i.e., water/sludge or certain lavas. Most of the channels can be traced back to Amphitrites Patera (likely one of Mars’ largest calderas) and large volumes of low-viscosity lavas are indicated by the area’s morphology. Although water/sludge flows remain a viable alternative to lava, previously proposed groundwater/-ice sapping [7] would not be expected in the hydrogeologically constrained setting of a caldera summit. An alternative is volcanically-induced melting of an ice sheet, which models [14] suggest to have accumulated on Amphitrites Patera in an ancient 1 bar atmosphere.

References
[1] Hynek et al. (2010). JGR: Planets, 115(E9), 1–14. [2] Alemanno et al. (2018). ESS, 5(10), 560–577. [3] Tanaka & Scott (1987). USGS IMAP 1802. [4] Tananka & Leonard (1995). JGR: Planets, 100(E3), 5407–5432. [5] Leonard & Tanaka (2001). USGS Geol. Inv., 2694, 80225. [6] Moore & Wilhelms (2001). Icarus, 154(2), 258–276. [7] Tanaka et al. (2002). GRL, 29(8), 1–4. [8] Bernhardt et al. (2016). Icarus, 264, 407–442. [9] Bernhardt & Williams (2019). Ann. M. of Plan. Geol. Mappers#7013. [10] Bernhardt et al. (2019). LPSC#1435. [11] Williams et al. (2009). PSS, 57(8–9), 895–916. [12] Willmes et al. (2012). PSS, 60(1), 199–206. [13] Bernhardt et al. (2016). JGR: Planets, 121(4), 714–738. [14] Fastook & Head (2015). PSS, 106, 82–98. [15] Peterson (1978). LPSC, 3411-3432. [16] Reyes & Christensen (1994). GRL, 21(10), 887–890. [17] Greeley et al. (2005). JGR, 110(E5), E05008. [18] Williams et al. (2005). JGR: Planets, 110(5), 1–13. [19] Williams et al. (2000). JGR: Planets, 105(E8), 20189–20205. [20] Elkins Tanton et al. (2001). Geology, 29(7), 631. [21] Arndt et al. (2008). Komatiite. ISBN 9780511535550. [22] Huppert & Sparks (1985). J. of Petrol., 26(3), 694–725. [23] Komatsu et al. (1992). GRL, 19(13), 1415–1418. [24] Whitford-Stark (1982). ESR, 18(2), 109–168. [25] Cattermole (1987). JGR: Solid Earth, 92(B4), E553–E560.