Design-Flood Flows in a Changing Climate

Bridges are designed to allow passage of flood flows to prevent their damage or destruction. Standard methods of computing design-flood flows, such as the 50-year flood, assume that flood flows (while variable from year to year) don't change over long periods of time. Annual peak flows on Maine rivers, however, have increased by a median of 18% during the last century. What will happen to flood flows in the next century if expected increases in temperatures and precipitation occur? It is unknown how well current Maine DOT design-flood methods will work in the future. We propose to test how well design-flood flows can be estimated for future scenarios. We have built and calibrated four detailed rainfall/runoff models for four watersheds in coastal Maine for a different project. We propose to use these models to hindcast historical annual peak flows, compute design-flood flows from the modeled peak flows, and then compare these design-flood flows to observed design-flood flows that are based on actual streamflow record. Following this, we propose to model future annual peak flows using multiple global-climate-model scenario projections, and compute future design-flood flows. We also will investigate model sensitivity by computing peak flows and design-flood flows for a specified set of climatic changes, for example, a 2 degree increase in air temperature or a 20% increase in precipitation.