Looming more than four kilometers above sea level, the Tibetan plateau is both the largest and highest plateau on Earth. Home to two of the world’s tallest peaks, Everest and K2, the “Roof of the World” owes its literally breath-taking topography to the geological equivalent of a car crash between the Indian and Eurasian plates — a collision that continues to this day, causing earthquakes and occasionally forcing a revision of Mt. Everest’s official altitude. 

But while scientists mostly agree on the broad strokes of Tibet’s chaotic formation, the details — specifically whether or not the plateau grew taller after its initial formation — have remained something of a geologic riddle. 

Reconstructions of plate movements and stratigraphic evidence suggest that it did, potentially shooting up abruptly between 20 and 10 million years ago and rising a bit more gradually after that. But when scientists used the Tibetan fossil and isotope records to reconstruct the paleolatitude of the region, they found no evidence of elevation gain. Indeed, according to fossil evidence, Tibet might even have sunk by up to a kilometer

Now, University of Copenhagen geologist Giampiero Iaffaldano thinks he’s found a way to reconcile the apparently conflicting evidence. 

To retrace the geologic past of the “Roof of the World,” Iaffaldano looked down to the ocean floor. Seafloor rock preserves snapshots of the Earth’s magnetic field captured at seafloor spreading centers — places where hot rock from the subsurface rises to become new seafloor. Those snapshots can be used to reconstruct the movement of the Earth’s crust in the distant past, and Iaffaldano used them to constrain India’s motion during Tibet’s formation. 

With a better idea of how plate motions unfolded over the last tens of millions of years, Iaffaldano was able to calculate how much force the Indian plate may have experienced as it rammed into Eurasia. By comparing his result to other scientists’ estimates of the amount of force needed for altitude increase, he determined that Tibet probably did rise — just not by very much. 

Critically, the modest altitude gain Iaffaldano suggests — just a few hundred meters — wouldn’t be enough for paleoaltimetry records to accurately resolve. If he’s right, both lines of apparently contradictory evidence for Tibet’s ancient altitude can be neatly accommodated into one unified story.