How We Learned to Distrust a Clean Number

There was a time, not long ago in the scale of a life now reaching past a hundred years, when a reassuring statistic reassured. A survey that reported ice loss "within model" would have settled the matter for a generation that trusted the model. It does not settle it now, and the reason is instructive. We have lived through enough clean numbers that later turned restless — the reservoir that was "adequate" until the summer it was not, the tolerance that held until it didn't — that the public has learned a hard literacy: read the footnote, weigh the word "within," ask who chose the model. The Polar Cryosphere Survey, which will file its quarterly Antarctic bulletin later this month, works in exactly this climate of earned suspicion, its careful "within-model" language heard by an audience trained to hear what such language has sometimes concealed. This is not paranoia. It is memory. A century that kept being surprised by its own graphs finally stopped trusting the smooth part of the curve — and started, wisely, reading the edges.