Date Awarded


Document Type


Degree Name

Doctor of Philosophy (Ph.D.)


Virginia Institute of Marine Science


Harry V. Wang


The barotropic and baroclinic responses of the Chesapeake Bay to forcings from two hurricanes were investigated by using unstructured-grid three-dimensional hydrodynamic models. The model domain includes Chesapeake Bay proper, the tributaries, and its extended continental shelf in the mid-Atlantic Bight. Two hurricanes were studied: Hurricane Floyd of September, 1999 and Hurricane Isabel of September, 2003, both of which made landfall within 100 km of the Chesapeake Bay mouth. Hurricane Floyd in 1999 passed through the entrance of the Bay from southwest to northeast along the coastlines of Virginia as a Category 1 storm, whereas Hurricane Isabel in 2003 made landfall on the east coast of North Carolina and moved inland toward the northwest as a Category 2 storm. For the barotropic simulation of the Bay responding to the hurricanes, the model results were compared with Bay-wide water level observations and the model showed reasonable prediction skill. It was found that the storm surge has two phases: a primary surge induced by the remote winds and a secondary surge induced by the local winds. For both hurricanes, the primary surge induced by remote winds propagated into the Bay initially, but the subsequent phase, influenced by the local wind, was notably different. Hurricane Floyd was followed by northerly (down-Bay) winds, that reduced the primary surge effect and caused a localized set-down; Hurricane Isabel was followed by southerly (up-Bay) winds, which superimposed on the primary surge effect and caused a localized set-up. The volume and salt fluxes were estimated at selected cross-sectional transects throughout the Bay, and it was found consistently for each transect that the net influx dominated during Hurricane Isabel while the net outflux dominated during Hurricane Floyd. For the Bay's tributaries, the large inland river discharge at the headwater can couple with the storm surge event to increase sea surface elevation on the second phase of sea surface elevation rise, which has a significant impact on inundation in the local low-lying areas. For the baroclinic response of the Bay to the hurricanes, the model results agreed reasonably well with additional observed data: sea surface elevation, velocity, and salinity profiles. From the perspective of salt flux, oceanic saltwater influx was evidently pushed into the Bay from the continental shelf at the initial phase of Hurricanes Floyd and Isabel associated with storm surge and salt intrusion. In the second phase, follow up with, down-bay local winds of eastern-type storms tend to enhance the stratification whereas up-Bay local winds of western-type storms tend to reduce the stratification. The hurricane surface wind stress is the primary agent for destratifying water column by transferring generated turbulent kinetic energy to the lower layer. The wind-induced straining during Hurricane Floyd was verified using non-dimensional parameters that incorporated the wind direction and the horizontal salinity gradient. Direct precipitation of hurricane rainfall acted more like a point source onto the Bay surface water, which created a layer of low surface salinity on the sea surface. It has an implication dynamically on generating a sea surface horizontal pressure gradient and re-distributing salinity field after the storm. Extra efforts have been made to conduct idealized experiments for comparing long-term recovery of the Bay to the disturbance created by the two hurricanes. Realistic hurricane wind forcing was applied in a 4-day window with the same initial condition applied in the beginning, and the quasi-steady state condition achieved in the end. Through this exercise, it was found that it took Bay 5-7 days to return to normal condition from the sea surface elevation disturbances for both Hurricanes Floyd and Isabel. For the salinity fields, it took within a range of 20-30 days to recover to the pre-storm condition for the middle and southern portions of the Bay. For the northern portion of the Bay, however, due to the landward barotropic pressure gradient generated a strong salinity rebound and the associated oscillation subsequently after Hurricane Floyd passed, it required twice as long to recover. Sensitivity testing of the effect of river discharge (immediately after the storm) on the recovery time has also been performed. Lastly, the influences of continental shelf dynamics on the Bay's response to the hurricane were examined. It was found that the along shelf wind contributed to the inflow and ouflow at the Bay mouth in the form of Ekman transport, which complemented the contribution generated by the Bay's local wind. The onshore and offshore shelf wind also played a significant role. Because the cyclonic pattern of the hurricane wind field, when the hurricane made the landfall in the US East coast, an along-the-shelf pressure gradient from the north to the south was generated. This pressure gradient, coupled with the Coriolis and friction forces, can generate a quasi-geostrophic balance flow serving to prevent or enhance the inflow across the Bay mouth. The effect is particularly noticeable in the relaxation period during the hurricane passage.



© The Author