Date of Award


Degree Type


Degree Name

Master of Science in Oceanography



First Advisor

Kathleen Donohue


The Kuroshio Current, a mid-latitude western boundary current in the North Pacific, is the main transporter of warm water northward. The current leaves the Japanese coast around 35°N, as the eastward flowing Kuroshio Extension (KE). This strong jet marks a front between the warm subtropical and cool subpolar waters. Meanders in the KE are the main mechanism for the transfer of heat and momentum across the front. The goal of this study is to quantify the frequency, wavelength, and growth of KE meanders in three state-of-the-art high-resolution general ocean circulation models. This study is timely due to recent analysis of two observational data sets: global satellite sea surface height which provides a basin-wide view of the KE and a regional process study that resolves the full-water column meander structure. These studies provide quantitative metrics that can be used to test each model's ability to reproduce the continuum of KE meanders. Recent analysis from the Kuroshio Extension System Study, KESS, quantifies the frequency, wavelength and growth of KE meanders between 143° and 149°E with periods 3-60 days. Comparable analysis is performed on output from the high-resolution global Hybrid Coordinate Ocean Model (HYCOM) circulation model and the Ocean General Circulation Model for the Earth Simulator (OFES) model. Two model simulations from HYCOM are considered, one with data assimilation ("Hindcast") and one without data-assimilation ("Simulated"). Metrics used to study the large-scale KE are path length, position and KE strength. Hindcast has the best fit for each metric compared to satellite altimetry as expected since this model assimilates sea surface height anomaly (SSHA) data. It reproduces the decrease in KE strength and increase in path length associated with KE system shift from stable to unstable regimes. OFES has a slightly more northern position compared to the observations, yet strength and path length are comparable to observations. Simulated has the mean KE position about 1.26° south of observed position and is weak compared to OFES, Hindcast and observations.

SSHA is decomposed into two components: variability from changes in the mass of the water column (termed deep) and steric contributions from changes in density and volume without changing mass (termed upper). In addition, SSHA is examined in three frequency bands: 60-30, 30-15 and 15-7 days to study meander-scale variability. In all models upper SSHA is strongly tied to meanders in the KE: peak upper variance occurs along the axis of the KE, and upper variance is strongest in the 60-30 day frequency band and decreases with increasing frequency. Variance in the upper SSHA is larger than the deep SSHA by almost a factor of 10. The deep-ocean SSHA variance suggests that deep variability arises due to a combination of instabilities in the upper jet, remotely-generated deep variability, and the ocean's response to atmospheric forcing. Simulated upper and deep SSHA variance is the weakest amongst the three models. Propagation speeds of upper KE meanders were estimated along the KE mean path and produced similar results to the observational studies. Phase speeds are fastest in high frequencies (30 kmd-1, 11 day period, 300 km wavelength) with decreasing speeds with longer periods (10 kmd-1, 45 day period, 500 km wavelength). KESS observations indicate that the interaction between KE meanders and external westward propagating deep anomalies are important for the local baroclinic intensification of KE upper meanders. Large-scale, likely atmospherically-driven, deep SSHA modes occasionally interact with eastward meanders in OFES and Simulated. Deep variability, originating at or near Shatsky Rise, occurs in all models and propagates west and interacts with eastward KE meanders. The phase relationship favorable for baroclinic intensification, deep leading upper, appears only in OFES and Simulated. It is hypothesized that the data assimilation scheme, which utilized only upper ocean data, employed in this version of assimilating HYCOM leads to the misalignment. Whether the upper-assimilation scheme or misrepresented deep propagation is responsible has yet to be analyzed. Further analysis of this version of Hindcast to study meander dynamics such as eddy heat or momentum fluxes is not recommended. Subsequent improvements to the assimilation scheme could easily be evaluated by repeated similar analyses to those presented here.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.