Mars’ north polar ice is far cleaner than the 25% dust scientists once estimated

Exposed water ice at the north pole of Mars holds about 3% dust by mass, according to a study from the University of Washington’s Applied Physics Laboratory, where earlier work had put the figure for the top layer as high as 25%. The paper in npj Space Exploration, published Sept. 8, 2026, states that the ice “generally contains less than 3% dust by mass.” The university announced the result on Sept. 29.

Aditya Khuller, a senior research scientist at the Applied Physics Laboratory’s Polar Science Center, framed the stakes in terms of heat, comparing dusty ice to a dark T-shirt in the sun: in the release he said dusty ice gets warmer and vaporizes faster on Mars. Dust content therefore bears on how quickly polar ice vaporizes.

Three percent against twenty-five

The university’s release states the old figure plainly: previous estimates suggested that the top layer of polar ice contained as much as 25% dust by mass, and the new work puts it closer to 3%. Khuller described the earlier situation as one of widespread disagreement: water ice around the north pole was not in doubt, he said, but how dusty that ice is had been contested. The disagreement is why a single new calculation, applied to several sites at once, carries weight with the polar-science community.

The paper itself, by Pari Mohan, now a recent University of Washington geoscience graduate, and Khuller, adds the spread behind the central number. Earlier studies ranged from below 0.1% dust (Kieffer, 1990) to as much as 15% (Langevin and colleagues, 2005), and the abstract cites previous estimates of up to about 25%. The new surface values sit mostly below 3%, roughly four times lower than Langevin’s at one location and about two orders of magnitude lower at Korolev Crater.

Six sites and a snow model borrowed from Earth

The method is the reason the numbers moved. The authors applied a physically based model built for terrestrial snow and ice, based on Steve Warren’s approach for snow and ice on Earth, after finding that the leading method, developed for lunar soil, looked inaccurate when checked against Earth data. They also tracked how the brightness of the surface changed over time.

Data came from six locations, as the paper lists them: two spots on the residual polar cap, Korolev crater, Louth crater, Chasma Boreale and the Phoenix landing site. The measurements draw on spectra from the European Space Agency’s Mars Express and from NASA’s Mars Reconnaissance Orbiter, plus surface imaging from NASA’s Phoenix lander, which touched down on May 25, 2008, at 68.22 degrees north, an arctic plain at a latitude comparable to northern Alaska.

Korolev crater shows the largest correction. Earlier Mars Express work there implied roughly 15% to 22% dust by mass, as Sci.News reports, while the new study finds roughly 0.2% to 0.4% in the top layer, a drop of about a factor of fifty at the low end of the old range. The Phoenix site, where estimates had varied by up to a factor of 25 between earlier studies, is now among the six locations handled with one method.

Layers like an ice-cream sandwich

The 3% figure describes exposed surface ice, not every layer in the cap. According to the university, the north pole is stacked like an ice-cream sandwich, with dustier ice layers between cleaner slabs. The paper quantifies the contrast: clean layers at 0.05% to 0.5% dust, interbedded with dust-rich marker beds at 25% to 75%. A 25% reading can therefore be a real number for a dusty bed while still being the wrong number for the surface ice, and the gap between the two is the practical lesson for anyone modeling this ice: the answer depends on which layer is being sampled, how deep the sample goes and what season the surface was observed in.

Seasons complicate the picture further.

Each winter a dusty frost forms and then disappears in summer, exposing older, cleaner ice beneath it. The paper reports fine-grained frost with a radius of roughly 60 to 100 micrometers early in the season, replaced later by coarser, cleaner firn of roughly 400 to 2,000 micrometers, a change the authors attribute to frost sublimation and exposure of underlying ice rather than grain growth in place. Earlier work had suggested grain radii of about 100 to 500 micrometers for the residual cap surface.

The layers also hold climate information. The university’s release says they preserve clues about Mars’s climate thousands of years ago, which only reads cleanly if the dust in each layer is measured correctly. The paper’s title, “Revised dust content and grain size of exposed water ice at the North Pole of Mars,” names both quantities it revises, and the six sites it covers now share one method.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

More from Morning Overview