Via Mike Campana, an interesting essay by Chris Milly about the problem climate change poses for the folks who build dams and such:

I would argue, in fact, that changing climate is the new “default hypothesis,” rapidly
displacing the assumption of stationarity upon which generations of hydrologists and engineers have built their careers—not to mention untold dollars worth of dams, wells, levees, reservoirs, hydroelectric power plants, bridges, irrigation systems, and culverts.
Stationarity is the assumption that the future will be similar to the past, in a statistical sense. Historical observations have been the rawmaterials for hydrologic analyses under the fast-fading regime of stationarity. If we can no longer invoke stationarity to convert observations into predictions, what can we do? What additional ingredients are needed for hydrologic analysis?