Riverbed ice is not glue: lab flume finds Arctic beds erode 10 times faster while thawing

The sign flip lives in the bed's permeability. In unfrozen gravel, water works its way down between the grains, and that loss into the bed takes energy out of the surface flow ^[7]. Freeze the bed solid and the pores shut, so the whole discharge stays on top and runs faster for it ^[8]. The stage that produces the result is between those two: enough of the surface has thawed for water to inject back into the sediment, but ice remains below, so the injection has something solid to push against, turns upward, and works grains loose from underneath ^[9].

The tenfold figure ^[1] is a ratio measured on glass beads, so read it as a demonstration that the mechanism exists rather than as a sediment yield. What travels better than the ratio is the timing. Take the ratio at face value and suppose the high-rate window occupies a tenth of the period in which a channel moves sediment: that window then carries roughly 53 percent of the season's total, more than the other nine tenths together ^[15]. The duration is my assumption, and duration is exactly what has not been measured here. The arithmetic only makes the structural point that a short window running an order of magnitude hot dominates an annual budget, so a survey that samples the warm middle of summer can be careful and still miss most of the work. Erosion that organises itself into steps and pools ^[10] compounds that, because it concentrates rather than spreading evenly along a reach.

The provenance is unusually plain for a counterintuitive result. It began with Shawn Chartrand's field observations in 2019 and became a Simon Fraser and UBC project that Jonas Eschenfelder picked up in summer 2024 ^[13][14]. Chartrand says he emailed colleagues who did not believe him, and the team ran a second experiment that returned the same outcome ^[4]. That is the right disclosure, and it also sets the standard the finding has to clear: the people best placed to reject it were the authors, they tried, and the next attempt has to come from someone else's channel.

The unresolved part is scale. Uniform non-cohesive spheres ^[5] are not a High Arctic bed of graded sand, gravel, silt and ice-cemented fines, and cohesion is the property most likely to change the answer, since it is the thing the old glue assumption ^[2] was really about. In a region warming about four times faster than the global average ^[12], the difference between a lab mechanism and a field flux is what any budget or scour calculation actually needs. The account supplied reports no field measurements and breaks off mid-sentence in its section on Arctic geopolitics ^[16].