Date of Award
2026
Degree Type
Dissertation
Degree Name
Doctor of Philosophy in Oceanography
Specialization
Physical Oceanography
Department
Oceanography
First Advisor
Lora Van Uffelen
Abstract
The Western Arctic Ocean is warming at an alarming rate. Massive loss of multi-year ice and a complete ice-free surface during the summertime allow for solar radiation and wind stress curl to impart heat and momentum into an otherwise quiet region. Subducted warm and salty water intrusion from the Pacific Ocean is further changing the water column stratification of the Beaufort Gyre, leading to a strengthened sound-speed waveguide structure, the Beaufort Duct. This duct is capable of propagating low-frequency sound over hundreds of kilometers with negligible loss. This dissertation focuses on in-situ observations collected by Autonomous Underwater Vehicles (AUVs) during an acoustic propagation experiment to study the impact that the gyre-modulated sound-speed structure and its fluctuations have on acoustic arrivals measured and used for underwater vehicle localization in the Beaufort Gyre region.
In Chapter 1, we assess the sound-speed fluctuations in the Beaufort Gyre region using observations collected by four AUVs as they navigated around and within the 2016-2017 Canada Basin Acoustic Propagation Experiment (CANAPE) acoustic array. The CANAPE array was deployed within the Beaufort Gyre for a full natural year and consists of a Distributed Vertical Line Array (DVLA) and six moored sources with receivers arranged in a 150-km radius pentagonal array with one moored source at its center. The AUV observations, which are of a spatial nature, were assumed to be taken synoptically, frozen time approximation, and are compared to previously published results of a temporal analysis from a DVLA which highlighted dominant spice-driven fluctuations in the upper 50-100 m. Sound-speed fluctuations were found to be primarily driven by warm-salty water intrusions (spice), in the upper 100 m of the water column followed by variability from internal waves and halocline eddies with effects down to 450 m. The results were successfully validated with the DVLA temporal analysis and fluctuations were reported for the first time up to the surface. Additionally, sound-speed profiles estimated with gliders in the Canada Basin show that spice-driven fluctuations double at the western edge of the Beaufort Gyre, highlighting the need to assess fluctuations and the impacts to sound propagation in the context of the gyre.
In Chapter 2, the focus is on analyzing the impact that a moving receiver’s Doppler effect has on acoustic ranging accuracy. The AUVs are moving receivers, in that they are continuously moving while receiving an acoustic signal, and as such are subject to Doppler frequency shifts. A shift in frequency gives the appearance that a signal is arriving earlier or later than it actually is. Here we leverage vehicle speed and attitude data to estimate the Doppler shift that linear modulated frequency signals sent by the CANAPE sources undergo as they are received by the moving AUV. Then an analysis focuses on the effects that this has on ranging estimation error as a function of source-receiver range. Typical AUV horizontal speeds of 0.25 m/s can lead to Doppler-induced range errors of 150 m if the receiver has radial speed components in the direction towards or away from a moored source. As the AUV moves closer to a moored source the contribution of this induced error to the total source-receiver range becomes larger than 10% of the total range. Results show that Doppler-induced errors are ubiquitous, due to the AUVs navigation paths which have them moving from one source to the next, and that they can become a dominant source of acoustic ranging uncertainty at short ranges.
In Chapter 3, findings from chapters 1 and 2 are used to achieve subsurface vehicle localization. Expanding on the results of chapter 1, an analysis of the measured acoustic arrivals compared to a predicted acoustic timefront highlights that ray paths with turning depths shallower than 500 m are difficult to predict accurately and are not ambiguously identified. These ducted arrivals are subject to a double-duct system, composed of the Beaufort duct and a seasonally strengthened surface duct, that induces reverse geometric dispersion on low order acoustic normal modes making them arrive earlier than higher-order modes. Additionally, ducted arrivals undergo scattering due to the sound-speed fluctuations present in the upper 450 m. These effects are tied to the Beaufort gyre’s depth of influence. Non-ducted arrivals, from ray paths with turning depths deeper than 500 m, spend most of their propagation path beneath the gyre’s influence and away from the upper ocean variability, making them predictable and identifiable across the Beaufort Gyre region. This makes the non-ducted or early arrivals ideal for acoustic ranging between source and receiver. These ranges are Doppler corrected following the methods laid out in chapter 2. Finally, a localization approach that leverages vehicle data is implemented to first localize a full reception group, 3 receptions or more, and second, acquire subsurface vehicle positions for each reception. Localization uncertainty estimates were found to be dependent on the AUV trajectory with larger values found south of the array and near moored sources where the localization constraints due to receiver and source geometry are ill conditioned. Overall, localization uncertainty is less than 100 m rms in x and y for all four AUVs. For reception groups with 4-5 receptions per localization the uncertainty is less than 63 m rms in x and y.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Pomales Velazquez, Luis Orlando, "GLIDER-BASED OBSERVATIONS OF GYRE-MODULATED SOUND-SPEED STRUCTURE IN THE CANADA BASIN: EFFECTS ON LONG-RANGE UNDERWATER ACOUSTIC PROPAGATION AND VEHICLE LOCALIZATION" (2026). Open Access Dissertations. Paper 4605.
https://digitalcommons.uri.edu/oa_diss/4605