Gentle Slopes, Fast Wear: Soft Rock Outpaces Its Harder Neighbour

Shale, a soft rock, is wearing away about three times faster than neighbouring hard rock in a Colorado mountain watershed, even though its slopes are gentler, according to a peer-reviewed study led by Eyal Marder of the University of Massachusetts Amherst ^[1][3].

The team also reports that the faster erosion increases the release of carbon dioxide from rock weathering ^[3]. That could matter beyond Colorado, though no one has yet measured by how much.

The paper, "Bedrock erodibility governs transient erosion and CO2 release in alpine landscapes," appeared online in Science Advances on October 14, 2026 ^[3]. The university published its own account on October 30 ^[1].

What the team measured

The study covers the East River watershed in Colorado, where shale sits alongside tougher intrusive igneous rock, the kind that forms when molten rock cools underground ^[1][3]. The researchers compared catchments dominated by each. The shale ones were being lowered about three times faster than the ones dominated by intrusive rock ^[3].

That runs against the usual expectation. Landscape theory says weak rock forms gentle terrain, while resistant rock can hold up steep, high ground. By that logic the steep, hard-rock catchments should be shedding material fastest. The gentle shale catchments are doing it instead ^[3].

The paper adds that such differences between rock types can persist for millions of years while a landscape adjusts, even where gentler terrain would suggest slower erosion ^[3]. Terrain alone, in other words, can hide where the fastest erosion is happening.

A clock inside quartz grains

The comparison was hard to make until now. Geologists estimate how fast a landscape is eroding by measuring beryllium-10, an isotope produced when cosmic rays strike quartz near Earth's surface. Material that has spent longer close to the surface has collected more of it, so the amount in river sediment helps reveal how quickly the ground is being stripped away ^[1][3].

The method has a weak spot. Shale is fine-grained, and its quartz grains are tiny and hard to isolate, so direct catchment-scale measurements in shale have been scarce ^[3]. Marder's team developed a new way of measuring beryllium-10 in shale. They paired it with conventional measurements in the intrusive rocks, then added topographic analysis, stream chemistry data and landscape-evolution modelling ^[3].

The study treats the watershed as a landscape still adjusting, and its carbon result applies to that period of adjustment ^[3].

Where the carbon comes in

Weathering is the chemical breakdown of rock by water. It releases dissolved material, called solutes, into streams, and some of the reactions involved exchange carbon dioxide with the atmosphere. Fast erosion keeps stripping off the weathered skin of a rock and exposing fresh material to the water.

The team's stream chemistry data and modelling indicate that faster shale erosion increases solute generation. They also indicate it increases net CO2 release from inorganic weathering while the landscape keeps adjusting ^[3]. The word "net" matters, because weathering reactions can pull carbon dioxide out of the air as well as put it in.

In this setting, the paper finds the balance tips toward release.

This is a finding about inorganic weathering. It does not show how much CO2 escapes when erosion exposes organic carbon, the remains of ancient living matter locked in some rocks ^[3]. That is a separate question the paper leaves alone.

How far the result reaches

The erosion contrast is established for this watershed. The size of any effect on the global carbon budget is not.

The study does not claim to have measured a worldwide emissions total, and it has not changed any climate projection. The accessible abstract also does not give the number of samples or the detailed uncertainties.

Marder said the team wants to understand the full geological history of a landscape, not only the snapshot visible today ^[1][2]. He also said that extending the work to mountain regions worldwide could improve understanding of biogeochemical cycles and climate change. In his account, such measurements could be used to estimate effects on CO2 emissions over thousands to millions of years ^[1][2].

That is a research programme, not yet a result, and its timescale is worth keeping in view. The release the paper describes plays out over thousands to millions of years. It is no explanation for the carbon dioxide rising in the air this decade, and no reason to doubt the burning of fossil fuels as the cause.

But it is a pointed reminder of how much the ground does. Carbon accounts of the planet usually treat rock as the slow, quiet background. Here, one rock type is shown to be wearing away three times faster than the hard rock beside it, in a place where the slopes suggested the opposite. If the same pattern holds in other ranges, models built on slope and steepness alone would be misplacing the erosion.

Whether it does hold is the open question. Mountain ranges differ in climate and in the mix of rocks, and the new beryllium-10 method has so far been shown to work in one watershed.

Other teams will need to apply it to shale in other catchments, and to find out what sampling it requires. Marder's group has said it wants to take the work to mountain regions around the world ^[1][2].