How high were these mountains?

An iterative approach is needed to constrain the past elevation of mountain belts Paleoaltimetry—the quantitative estimate of the past elevation of land surfaces such as mountain belts—is notoriously difficult to constrain. To estimate past elevation, geologists study sedimentary rocks that accumulated in freshwater lakes in ancient mountain belts. They compare fossils or oxygen stable isotopes (the ratio of which is elevation-dependent) from these rocks with present-day records from elevated areas ( 1 ). One region where paleoaltimetry studies are widely conducted is the Tibetan Plateau, which owes its extreme elevation to intense deformation that started at least 70 million years ago during subduction of the Indian plate below Asia. However, elevation estimates for the region—for example, for the Eocene (∼40 million years ago)—based on either fossils or stable isotopes differ by kilometers ( 2 , 3 ). On page 946 of this issue, Botsyun et al. ( 4 ) provide an explanation for this discrepancy and bring the two types of estimate nearer to agreement.