About 18,000 years ago, sea level sat roughly 100 meters below where it is now. The ice melted, the channels opened, and the water moved fast enough to scour bedrock at Niagara Falls velocity. The Yorath Island Spillway event, dated to around 12,000 years ago, deposited Canadian Shield sediments hundreds of kilometers south in a single catastrophic discharge. That was not a slow process. It was a pulse.
The question worth asking in 2026 is whether the plumbing that drove those ancient pulses is active again, and whether our sea level models account for it honestly. The answer to the second question is almost certainly no.
What the Channels Actually Do
Subglacial meltwater channels do not just drain water. They reduce friction between the ice sheet base and the bedrock beneath it. Less friction means faster ice movement toward the ocean. Faster ice movement means more calving, more mass loss, and more sea level rise. The physics is not complicated. The measurement problem is.
We do not have a real-time network monitoring subglacial channel activity under the West Antarctic Ice Sheet or the Greenland Ice Sheet at the resolution needed to quantify this effect. What we have are satellite surface velocity measurements, ice-penetrating radar surveys with significant gaps, and paleoclimate records showing what happened when these systems ran at full capacity. The paleoclimate record is alarming. Sea level reached 6 meters above current levels roughly 125,000 years ago, during a period when global temperatures were only modestly warmer than today.
Isostatic rebound complicates the picture further. Eastern Hudson Bay is still rising at 1.3 cm per year as the crust recovers from ice that melted millennia ago. Near active ice sheets, that same rebound effect masks how much the bedrock is actually subsiding under current ice load, which distorts local sea level measurements. The signal we most need to read clearly is the one most obscured by geological noise.
The Planning Problem Nobody Wants to Price
Fair point to the skeptics: the research brief for this piece finds no recent study confirming that subglacial channels are accelerating sea level rise faster than current IPCC projections assume. That absence of confirmation matters. I am not claiming the science is settled against the models.
But the absence of confirmation is not the same as evidence of absence. The channels existed. The floods happened. The mechanism is physically real. What we lack is the instrumentation to measure whether it is happening now at a rate that should change the numbers coastal engineers are using.
Those engineers are currently designing seawalls, drainage systems, and flood barriers for cities like Miami, Jakarta, and Rotterdam using projections that may not fully account for subglacial lubrication dynamics. Miami-Dade County alone has committed roughly $4 billion to sea level adaptation infrastructure. If the baseline projection is low by even 20 centimeters by 2100, significant portions of that investment are undersized.
The fix is not another modeling paper. It is a funded, international subglacial monitoring network, the kind of instrumentation commitment that matches the scale of what we are trying to measure. NASA's Operation IceBridge produced useful data before it was wound down. Its successor programs are underfunded relative to the question they are trying to answer.
The ancient floods left their evidence in scoured bedrock and displaced sediment. We found that evidence thousands of years later. We do not have thousands of years to find the evidence this time.
