Credit: NASA / JPL / ASO

Identifying past oceans and lakes on Earth is — relatively — easy. Cliffs and road cuts reveal rocks rife with fossils and minerals formed in water. Seismic data reveal buried shorelines and channels. And radioactive elements allow us to assess when these markers were deposited. Finding former bodies of water on other planets is markedly more challenging. Spectral data can hint at the presence of aqueous minerals, but most evidence comes from satellite images of planetary landforms that resemble fluvial or marine features found on Earth.

The presence of such analogous surface morphology has led to the suggestion that the northern plains of Mars were once covered by vast oceans. Deep chasms in the Chryse Planitia and Valles Marineris regions have been interpreted as outburst channels that fed ground water or molten ice into the ocean basin. But these features may not be signs of a shoreline at all: Lorena Moscardelli and Lesli Wood of the University of Texas, Austin, suggest that elongated mounds near the chasms are more consistent with channel formation in a continental slope setting (Geology 39, 699–702; 2011).

The martian channels look like flood remnants, but also bear a striking resemblance to submarine channels found on Earth. Dotted among these channels are triangular elevated features termed teardrop-shaped islands. Moscardelli and Wood show that these martian landforms are geometrically similar to erosional shadow remnants found on the continental shelf near Trinidad. The Caribbean erosional remnants formed during a mass wasting event in the distal reaches of the Orinoco Delta, and similar processes could have been at work in the martian plains.

Traces of mass wasting events are present elsewhere in the northern martian plains, namely in the form of chaotic terrains in the upper limbs of the outflow channels near Chryse Planitia. This surface morphology is consistent with channel collapse and underwater transport of sediments.

If the channels were indeed carved under water, we may need to redraw the outlines of the hypothesized ancient martian ocean, and raise estimates of the volume of water that once occupied it.