Date of Award

2015

Degree Type

Thesis

Degree Name

Master of Science in Oceanography

Department

Oceanography

First Advisor

Kathleen Donohue

Abstract

Accurate high-resolution ocean models are required for hurricane and oil spill pathway predictions, and to enhance the dynamical understanding of circulation dynamics. In order to investigate Loop Current dynamics, including eddy-shedding mechanisms and the forcing of deep flow in the Gulf of Mexico, a mapping array centered near 26°N 87°W was deployed from April 2009 through November 2011. The array provides a unique dataset for studying the Loop Current eddy cycle: it was centered in the region of Loop Current eddy formation/separation and during its 30-month deployment observed four Loop Current eddy events with measurements throughout the water column at daily temporal and mesoscale spatial resolution. This dataset provides the critical deep-velocity information required for comprehensive model-data intercomparision, a necessary first step in order to use a model for dynamical interpretation. The 1/25° data-assimilating Gulf of Mexico HYbrid Coordinate Ocean Model (HYCOM31.0) represents one of the highest resolution data-assimilating simulations of the full Gulf of Mexico region. This study compares output from HYCOM31.0 to the array observations to assess HYCOM31.0's viability for use in studies of Loop Current processes, focusing on Loop Current path variability and upper-deep layer coupling during eddy separation. Point-to-point array averaged correlation was 0.93 for sea surface height (SSH), and 0.93, 0.63, and 0.75 in the thermocline for temperature, zonal, and meridional velocity, respectively. Peaks in modeled eddy kinetic energy during eddy separations were consistent with observations, but modeled deep eddy kinetic energy was half the observed amplitude. Loop Current meander phase speeds and wavenumbers, and site-to-site SSH coherence indicate high model accuracy, particularly for periods longer than 20 days. The model reproduced observed patterns indicative of baroclinic instability, that is a vertical offset with deep stream function leading upper stream function in the along-stream direction. While modeled deep eddies differed slightly spatially and temporally, the joint development of an upper-ocean meander along the eastern side of the Loop Current and the successive propagation of upper-deep cyclone/anticylone pairs that precede separation were contained within the model solution. Overall, the model-observation intercomparison indicated that the 1/25° Gulf of Mexico HYCOM is well suited for the study of Loop Current eddy formation and separation, offering a larger spatial and temporal window than observational arrays.

Share

COinS