2015 GSA Annual Meeting in Baltimore, Maryland, USA (1-4 November 2015)

Paper No. 134-15
Presentation Time: 9:00 AM-6:30 PM


STROMER, Zachary D.1, WOODRUFF, Jonathan D.1 and DONNELLY, Jeffrey P.2, (1)Department of Geosciences, University of Massachusetts Amherst, 611 North Pleasant St, 233 Morrill Science Center, Amherst, MA 01003, (2)Geology & Geophysics Department, Woods Hole Oceanographic Institution, MS #22, 266 Woods Hole Rd, Woods Hole, MA 02543, zstromer@geo.umass.edu

Prior to Hurricane Sandy, the return period for a flood of that magnitude in New York City, calculated using instrumental tide gage data was 10,000 years. It is now becoming clear that the true return period was likely closer to 100 years. This two-order of magnitude underassessment of risk for New York City highlights the inability of short instrumental data sets in estimating the risk of extreme flooding to the region. Hurricane Sandy now provides a modern and more accurate representation of flood risk for the New York City region. However, such an event is still not available for locations further to the north where Hurricane Sandy had less of an impact. The Great Colonial Hurricane of 1635, considered by many to be the historical event of record for many locations in New England, may provide such an event. Early colonial records provide documentation for storm tides as high as 6 m in Buzzards Bay and 4 m in Narragansett Bay during the storm. Historical accounts of the 1635 event may be prone to exaggeration, and significant questions have been raised regarding the occurrence and true intensity of this legendary storm. To provide further insight we present a large number of geographically spaced sedimentologically derived records combined with NOAA’s SLOSH (Sea, Lake, and Overland Surge from Hurricanes) storm surge model to estimate the maximum flooding surge at locations along the Massachusetts coastline. By adding this modeled storm surge data to commonly used flood return period models which use Generalized Extreme Value (GEV) theory, we show that GEV may be greatly over-estimating the return period (i.e. under-assessing risk), for larger flood events on the southern facing coastline of Massachusetts. Results emphasize the inability of short instrumental data sets for assessing flood risk by 100 yr or greater events, as well as the value in combining sedimentological, modeled and historical records of early historical floods for improving these assessments.