For more than 30 years, scientists have argued over whether a vast ocean once spread across the northern third of Mars. Now a team led by Abdallah Zaki of the University of Texas at Austin has found what may be the most durable clue yet: a broad, gently sloping band of terrain hugging the boundary between the planet’s low northern plains and rugged southern highlands that, by every measurable standard the researchers applied, matches the continental shelves that ring Earth’s ocean basins.
The findings, published in Nature in spring 2026, shift the debate away from hunting for ancient shorelines, which erode and warp beyond recognition over billions of years, and toward a feature that is far harder to erase.
A shelf instead of a shoreline
Previous searches for a Martian ocean focused on tracing thin, shore-like ridges along the edge of the northern lowlands. The idea dates to a 1993 paper by Timothy Parker and colleagues, who spotted what looked like coastal landforms in spacecraft images. But critics pointed out a stubborn problem: the candidate shorelines do not sit at a single, level elevation, as they should if they once marked the rim of a standing body of water.
Zaki’s team took a different approach. Instead of looking for a narrow line, they searched for a wide, flat apron of terrain, the kind of gentle slope that forms where an ocean meets a continent and sediment slowly accumulates over long stretches of time. On Earth, that feature is the continental shelf, and it can persist in the topographic record long after any waterline has vanished.
To make the comparison rigorous, the researchers benchmarked their Mars measurements against the Global Seafloor Geomorphic Features Map, a peer-reviewed digital atlas that classifies Earth’s ocean floor into distinct zones including shelf, slope, abyss, and trench. They applied the same slope-based criteria to elevation grids of Mars built from NASA’s Mars Orbiter Laser Altimeter (MOLA), which has mapped the planet’s surface to a vertical precision of roughly one meter. The Martian band, they found, falls within the same range of gradients and widths as terrestrial shelves cataloged in the atlas.
“Continental shelves are more stable ocean fingerprints than fragile shorelines,” Zaki and coauthor Michael Lamb of Caltech explained in an institutional release from UT Austin’s Jackson School of Geosciences. A separate summary from Caltech described the feature using a “bathtub ring” analogy: the shelf-like zone implies a stable, long-duration ocean rather than a body of water that appeared and vanished quickly. According to the UT Austin release, the ocean may have persisted for “possibly millions of years.”
Why the method matters
Because the analysis draws on global topography rather than any single crater rim or valley outlet, it does not depend on one landform being correctly interpreted. The method searches for a continuous band of low slopes that follows Mars’ crustal dichotomy boundary, the sharp geological divide between the smooth north and the cratered south, much as Earth’s continental shelves follow the edges of the continents. That makes the result harder to dismiss as a trick of local geology.
The quantitative, cross-planetary comparison also sets the work apart from earlier studies that relied on tracing visible features by hand. The study does not merely assert that a Martian slope “looks like” a shelf; it tests whether the slope meets the same statistical criteria that define shelves on Earth. For a hypothesis that has survived three decades of scrutiny without being conclusively confirmed or refuted, that methodological step is significant.
What the study cannot yet prove
Topography alone cannot confirm that water, rather than some other process, carved the low-gradient band. Volcanic resurfacing, ice-related subsidence, or long-term crustal warping could, in principle, produce a similar slope profile. No rover or lander has directly sampled sediments from the identified shelf zone, so the composition and layering of the terrain remain unknown.
A 2024 study in Scientific Reports used regional mapping around China’s Zhurong rover landing site in Utopia Planitia to examine the Deuteronilus contact, one of the proposed ancient shorelines, and linked it to units of the Vastitas Borealis Formation. That work documented elevation variability along the contact consistent with either crustal deformation or multiple water levels, but it did not test the broader shelf criterion introduced in the new Nature paper. The two studies are complementary rather than competing.
Timing and climate pose additional puzzles. A shelf that took millions of years to form requires liquid water stable at the surface, which in turn demands a thicker atmosphere and warmer temperatures than Mars has today. Current climate models struggle to sustain those conditions under the faint young Sun that illuminated the early solar system. Reconciling the apparent duration of the inferred ocean with those models will require better constraints on Mars’ early greenhouse gas inventory, volcanic outgassing rates, and the pace at which the atmosphere was stripped away by solar wind.
A 2003 review by Michael Carr and James Head, published through the U.S. Geological Survey, remains a standard reference on the evidence for and against ancient Martian oceans. But it predates major advances in orbital radar sounding, mineralogical mapping by ESA’s Mars Express and NASA’s Mars Reconnaissance Orbiter, and surface observations by multiple rovers. How the new shelf criterion fits into updated global models of Mars’ water budget is a question the authors acknowledge but have not yet addressed in detail.
It is also unclear whether the inferred ocean existed as a single, continuous body or as a series of episodes separated by freezing or partial drainage. Either scenario would leave a shelf-like imprint, but distinguishing between them matters for understanding how long Mars remained habitable.
What would settle the question
The shelf finding gives future missions a concrete target. If orbital radar surveys, such as those planned for ESA’s ExoMars program, detect sedimentary layering within the identified band, with grain sizes and structures characteristic of shallow-marine environments, the ocean interpretation would gain powerful support. Surface sampling by a rover or, eventually, returned samples analyzed in Earth laboratories could test for minerals that form only in sustained contact with liquid water.
Conversely, if instruments on the ground reveal predominantly volcanic or ice-modified deposits with no marine signatures, the shelf-like gradient would need a different explanation, perhaps differential crustal loading or tectonic processes unique to Mars. Either result would sharpen the picture of what early Mars looked like and whether its surface ever stayed wet long enough for life to gain a foothold.
For now, the mapped band behaves, in every statistical and geomorphic measure the researchers applied, like a continental shelf. That does not prove an ocean existed, but it makes the case more coherent and more testable than it has been at any point since the debate began in 1993.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
