Peter Knuepfer (SUNY Binghamton)
Evidence for Changing Flood Frequency, Upper Susquehanna River Basin
Major floods in central and southern New York in 2006 and 2011 were the largest of record at many locations. Estimates of the recurrence interval for these events ranged from greater than 100 years to greater than 500 years. These floods beg the question, were these events simply bad luck with the “law of averages” catching up, or do they represent a change in the frequency of severe flooding as a regional response to the changing climate? This question is examined by evaluating changes over time in the size of long recurrence-interval floods (1% and .2% annual likelihood events) predicted by conventional flood-frequency analysis from the Susquehanna River basin. I apply a 50-year moving window to long-duration gauge records, calculate the expected 1% and 0.2% for each window, and average the result for each year to develop an annual prediction curve for 100-year and 500-year floods. Most gauge stations show a sharp increase in the predicted magnitude of long recurrence-interval floods over the last few decades compared to the early-to-middle 20th century. This increase in the size of predicted floods is strong evidence that the recent floods indicate a change in potential for major floods in the upper Susquehanna basin, consistent with predictions that have been made for the effects of climate change. This also implies that static flood-hazard analyses may be underestimating future flood potential, even if they have been recomputed to include the recent floods.