RECURRENCE INTERVAL OF THE BINGHAMTON, NY, FLOOD OF JUNE 2006

Binghamton, NY, and surrounding areas experienced a flood of historic proportions June 27-July 1, 2006 due to heavy rains (up to 13-15 inches over 4 days in some locations) superimposed on moderately saturated soils. All gaged sites along the upper Susquehanna River (and some of its tributaries) upstream of Owego, NY reached record heights on June 28. Some 15,000 people were evacuated in the Binghamton area alone, and total flood damage will greatly exceed a quarter billion dollars. Key to understanding the impact of a large flood like this, and to planning for future floods, is assessment of the frequency of the event. Flood frequency analysis generally involves extrapolation of historic annual series (maximum annual stage or discharge) to longer time periods of interest, such as 100 years or more, assuming stationarity of the flood series. Outlier events–like the June 2006 flood–are excluded from the analysis, as they would skew the results. A Pearson Type III distribution is assumed for flood recurrence, and this kind of fit usually provides robust flood estimates. But the occurrence of major events generally requires a re-evaluation of the statistics. Preliminary analysis of the 2006 flood indicates that it exceeded an expected 500-year recurrence interval (0.002 annual probability) size at stations on the Susquehanna River upstream of Binghamton and on the upper Chenango River. However, the flood peak reduced in the downstream direction, with an estimated recurrence of around 250 years at Vestal and less than 50 years at Waverly, NY. This reduction in flood peak was caused by the reduction in contributions from tributary streams as one moved downstream from the Susquehanna-Chenango River junction in Binghamton. If one includes the 2006 and other large historic floods in the frequency analysis, the recurrence intervals reduce to 300 years at Conklin and 80 years at Vestal–still unusual floods, but frequent enough to give pause.