Implementation of a one-dimensional mixed layer model for predicting ocean thermal profiles

Thumbnail Image
Issue Date
Osburn, A. D.
Earth sciences
The purpose of this thesis was to select, implement, and validate an open ocean numerical mixed layer (ML) model for the purpose of forecasting the evolution of the upper ocean thermal structure. The selected ML model would replace a manual prediction method used by the Canadian Forces Meteorology and Oceanography Centre (METOC). An examination of applicable ML numerical models was conducted and the model chosen was an integrated ML model originally developed by Kraus and Turner (1967) and expanded upon by Denman (1973). The implementation of the Denman (1973) model, called the mixed layer evolution model (MLEM), was initialised with a thermal profile, mixed layer depth and temperature, and was forced with hourly meteorological data. Model sensitivity test cases originally conducted by Denman (1973) were repeated to ensure that the MLEM adequately represented the original Denman model. Bathythermograph and meteorological data from Ocean Weather Station (OWS) "Papa" in the period 18 May 1970 to 28 June 1970 were used for validating MLEM predictions and these were shown to follow both ML depth and temperature evolution well. MLEM temperature profiles were used as inputs to acoustic propagation loss models to gauge the potential acoustic impact of employing modelled temperature profiles. The acoustic study employed the SUPERSNAP1 normal mode theory acoustic model for low frequency runs, and the SAFARI fast field theory acoustic model for high frequency runs. The acoustic study revealed that MLEM predictions worked equally as well and often better than manual method predictions for acoustic propagation loss modelling.
This work was digitally reproduced from a non-circulating print copy held by Royal Roads University Library. It forms part of a limited-scope digital collection of locally significant historical theses for which the Library is not currently accepting requests for digitization or deposit. Please contact the Royal Roads University Library for more details. The author has granted the Royal Roads University Library the non-exclusive right to digitize and make this work electronically available via DSpace @ RRU. This work should not be copied, downloaded, or distributed further without permission from the author. Please contact the RRU Copyright Office for more information.
Harmful Language Statement