Many flood risk assessments are based on instrumental records less than a century in length. Sedimentary and historical archives provide the opportunity to extend flood records by several centuries to millennia. In doing so these longer flood records provide opportunities for improving upon current flood risk assessment and gaining additional insight on the various climatic and geomorphic processes that drive changes in flood frequency. Such a reconstruction has not been attempted previously for Boston, MA where extratropical cyclones (ETC) are currently the dominant mechanism of coastal flooding. Here, we present both a historical reconstruction of extreme storm tides to affect Boston Harbor, and an independent geologic assessment of extreme flooding based on flood deposits preserved within the sediments of Bartlett Pond, a back-barrier coastal pond located 60 kilometers south of Boston. The historical reconstruction presented here identifies events of extreme flooding back to 1723, which are temporally consistent with overwash deposits identified from the sedimentary analysis at Bartlett Pond. Bartlett Pond is beyond the influence of significant dredging, landfill and dam projects within Boston Harbor. The consistency between extreme flood occurrences in Boston and Bartlett Pond therefore suggest that these man-made alterations have had a minimal impact on extreme flooding to the harbor. Additional modeling work is necessary however to provide confirmation on this initial finding. While flooding associated with the Blizzard of 1978 appears to be an anomaly in the modern instrumental record, our new historical/sedimentological record identify 6 additional events of similar size since European colonization, suggesting an under-assessment of the risk of these types of extreme events for Boston by as much as 300%. Additionally, the 1000 yr Bartlett Pond sedimentary reconstruction appears to show an increase in overwash frequency over the last 300 – 500 years when compared to the 500 years prior. A similar increase in ETC flooding has been observed in nearby sedimentological archives from the Gulf of Maine and could possibly be explained by variations in the North Atlantic Oscillation (NAO) and/or changes in sea surface temperature.

The thesis comprises two separate journal articles that together form a coherent body of work. In this thesis, the key physical processes responsible for the 23-26 September 2012 UK floods are investigated using a case study approach. The cyclone responsible for the floods developed near the Azores on 20¬-22 September following the interaction between an equatorward-moving potential vorticity (PV) streamer and tropical storm Nadine. Convectively-driven latent heat release associated with the developing cyclone reduced upper-level PV and resulted in the fracture of the PV streamer into a discrete anomaly as the cyclone intensified. In Paper 1, convection-permitting model simulations and diabatic heating rate and PV tendency calculations along trajectories demonstrate that deposition heating strongly reduced upper-level PV in the vicinity of the PV streamer, contributing to its fracture into a discrete anomaly. The cyclone deepened further over the UK on 23-26 September, ahead of a second upper-level PV anomaly. In Paper 2, sensitivity simulations of the storm are presented. PV inversion is used to modify the strength and position of the PV anomaly in the initial conditions and to examine whether the event could have been even more extreme with different upper-level forcing. Results show that quasigeostrophic forcing for ascent ahead of the PV anomaly contributed to the maintenance of the rainfall band over the UK. Counterintuitively however, strengthening the upper-level forcing produced a shallower cyclone with lower rainfall totals. Instead of moving eastward over the UK to interact with the cyclone, the strengthened anomaly rotated cyclonically around a large-scale trough over Iceland, resulting in a fragmented rainfall band. The counterintuitive results suggest that the verifying analysis represents almost the highest-impact scenario possible for this flooding event.