Big Muddy Speaker Series presents:

Original Presentation by Bob Criss, PhD, Professor at Washington University

Wednesday, September 11, 2013

How many so called "100-year floods" have you witnessed in your lifetime? Why are we continually surprised when floods of "unprecedented" heights hit our riverside communities? What did we learn from the Flood of 1993, 20 years ago, and what did we ignore?

Recent floods on major Midwestern rivers have set all-time record levels at numerous sites, prompting great overuse of the terms “100-year”, “200-year” and “500-year” floods.  This overuse is not trivial because the posited “100-year” flood levels have regulatory significance and affect property values.  Statistical methods prove that the calculated flood levels for extreme events are much too low, particularly because population stationarity has been assumed in official calculations, even though many factors such as climate change and in-channel structures are causing flood levels to rise.  New methods are proposed for the estimation of future flood levels that take secular changes into account.  These methods show that the official “100-year” levels are too low by 1 to 3 feet at many locations, and suggest that average levels attained by extreme floods will continue to rise by about an inch per year.

Dr. Bob Criss of Washington University discussed the science behind forecasting (or mis-forecasting) flood risk as well as the effects of changes we are making to the floodplains affect flood risk.

This photo of an industrial park in the Chesterfield Bottoms in the flood of 1993. Photo from a St. Louis Post-Dispatch retrospective article.