| The primary objective of this dissertation is to describe the tidal and subtidal flow patterns over the inner shelf of the Delmarva Peninsula, located in the Mid-Atlantic Bight of the United States (36.6--38.0 N), north of the Chesapeake Bay. The objective is pursued with a combination of direct measurements and numerical assimilative techniques. The dynamic balance of the study area is little known, and the distribution of tidal properties has not been described for this area since very rough descriptions in the 1950's. Hydrographic and current velocity profiles from four regional cruises in the inner shelf were used to study the area. The tidal and subtidal fields were studied using data assimilation techniques on a numerical model. The model described the spatial and temporal dynamics of the area and included vertically averaged shallow water equations. Current velocity measurements were assimilated into the model using the adjoint method. Concurrent predicted sea level data from inside the Chesapeake Bay were also assimilated in order to incorporate the sea level signal in the model. Measured current velocities were not able to represent adequately the tidal signal in the location of sea level stations, except for one cruise. In turn, sea level data were not able to recover shipboard current measurements. A weighted combination of both data sources and a regularization term that penalized vorticity, gave the best results in terms of minimizing the root mean square error of un-assimilated information. The mean circulation obtained over the inner shelf was less than 10 cm s-1 and oriented along shelf. The mean flow and elevation reflected semigeostrophic dynamics with along shore pressure gradient balanced by friction and rotation, and cross shore pressure gradient balanced by rotation. The mean flow and elevation had spatial scales of 15--40km in the along shelf direction. The across shelf direction presented smaller scales (3--5 km). In terms of tidal flows, the semidiurnal constituent was dominant, with magnitudes of 30 cm s-1. The diurnal constituent was less than 10 cm s-1. The propagation of the semidiurnal tide could be explained as combination of a Kelvin and a Poincare wave that transform into a coastal trapped Kelvin wave as it moves into the Chesapeake Bay. |
