Every pipe, culvert, and flood channel in America was sized by feeding historical rainfall records into a formula. Decades of local weather data go in. Pipe dimensions come out. The underlying logic is simple and, until recently, sound: the assumption is that historical trends will continue into the future. With climate change, that assumption doesn't hold.
Peer-reviewed research shows stationary models can underestimate extreme precipitation by up to 60 percent. Engineers know this now. But the concrete is already in the ground, shaped to contain a past that no longer describes the future.
