Large volumes of water are hypothesized to have carved and passed through the Martian outflow channels (e.g., Baker, 2001). Because these channels originate from discrete sources, a groundwater origin is typically invoked (e.g., Head et al., 2003). Given the large discharges needed to create the observed landforms, in some cases a couple orders of magnitude greater than the largest catastrophic floods on Earth (Baker, 1982), large and permeable aquifers would be needed (e.g., Carr, 1979;Manga, 2004). While most of the outflow channels are Hesperian (e.g., Tanaka, 1997), their formation continued through the Amazonian (e.g., Rodriguez et al., 2015). Some of the youngest channels originated from fissures in Athabasca Valles, Eastern Elysium Planitia within the past 10s of millions of years (Burr et al., 2002;Voigt & Hamilton, 2018). The subsurface of Mars thus appears to have hosted and episodically released large volumes of water over most of Martian history. Hence, detecting the presence and quantifying the volume of subsurface water and ice would help constrain the water budget and cycle from the Noachian to present (e.g., Clifford and Parker, 2001), the amount of water lost to space (e.g., Jakosky, 2021), the fate of possible oceans (e.g., Citron et al., 2018), and the amount of water sequestered in minerals (e.g., Scheller et al., 2021).To discharge water at the surface, aquifers must have sufficient pressure for water to reach the surface. One way to achieve hydraulic heads greater than hydrostatic and hence enable surface discharge is to confine aquifers beneath an overlying ice-saturated crust or cryosphere (e.g., Andrews-Hanna and Phillips, 2007;Carr, 1996;Harrison & Grimm, 2004). As Mars cools and this cryosphere thickens, hydraulic heads will increase and may also create the pressure needed to fracture the crust (Wang et al., 2006). The MARSIS radar system has identified reflections interpreted as lakes confined under Mars' southern ice cap (Lauro et al., 2021;Orosei et al., 2018), though this interpretation is contested (Ojha et al., 2021) and would require recent magmatism (Sori & Bramson, 2019). Aquifers in the crust, if they exist, would be at depths of several kilometers (Clifford et al., 2010), deeper than the MARSIS and SHARAD radar systems can penetrate in volcanic terrain (e.g., Abotalib & Heggy, 2019). MARSIS has not identified a deep reflector indicative of an aquifer below the outflow channels in Athabasca Valles, for example (Clifford et al., 2010). Other geophysical data, such as seismic shear wave velocity, Vs, may be useful because Vs is sensitive to physical properties of the subsurface and probes greater depths.Our objective is to interpret Vs measured by the InSight mission in Elysium Planitia. We use rock physics models to compute effective medium properties and to help distinguish between porous basalt filled with gas, liquid water, ice, or mineral cement. We focus on two observations. First, Vs within Mars' upper