Onboard Wave Forecasts

Caires, Sofia Isilda Nunes de (1997) Onboard Wave Forecasts. MSc(R) thesis, University of Glasgow.

Full text available as:
[thumbnail of 11007669.pdf] PDF
Download (4MB)


The aim of this thesis is to provide an easy, fast and reliable way of forecasting the wave conditions onboard a ship once the wind fields are provided. The resulting methodology, called BOWFOR, is based on the assimilation of the second generation wave model DOLPHIN with wave observations onboard the ship. Specifically, BOWFOR is intended as a methodology for providing wave forecasts at any time-space location or set of locations chosen, and in particular along the ship route. It is executed by running a model on a personal computer, and requires wind field forecasts for the trip at the departure of the ship. The forecasts are produced and updated during the journey through the assimilation of DOLPHIN with buoy, satellite or visual observations. In the latter case, the observations are corrected by currently used empirical formulas. It will be seen that the methodology requires only four daily wave observations to produce good forecasts, and the forecasts are valid for at least two days. In the important situation in which communications are lost, the possibility of assimilating visual observations allows efficient forecasts. Thus the BOWFOR methodology, by regularly providing the captain of the ship with updated information on the wave conditions ahead, permits an optimised route planning. Setting up BOWFOR involved essentially the choice of an appropriate wave model and an efficient assimilation technique. It also involved the study of the wave theories behind wave models and of assimilation techniques, the adaptation of WAM and DOLPHIN, the wave models chosen for our investigation, the testing of models, analysis of results, and getting sensibility about the relevant variables and parameters, and the choice and acquisition of data and planning of the experiments. The thesis is divided into seven chapters. Chapter 1 is introductory. Chapter 2 describes the basics of linear wave theory, and gives an account of the representation of random sea waves by spectral analysis methods. Chapter 3 outlines the history of wind waves theory along with the physics behind wave models. In Chapter 4 the performances of WAM, a third generation wave model, and of DOLPHIN, a second generation model, are compared. The comparison is made through hindcasts of storms that occurred in the Portuguese coast. The chapter includes some details about the adaptation of the DOLPHIN and WAM models. Both models were run on a personal computer, as a way of providing straightforward comparisons in a realistic environment. It is shown that WAM is to a certain extent more accurate than DOLPHIN, but that it cannot be efficiently run on board ships. The relatively good results provided by DOLPHIN justify its choice as the model to be used by BOWFOR. Chapter 5 is dedicated to the improvement of DOLPHIN forecasts by a recently introduced assimilation technique. The technique is tested with observations (measured and simulated) made with buoys at the forecast location and upwave of the forecast location, and with satellite. It is shown that the continuous assimilation of significant wave height and mean wave period is very efficient, providing even better results than WAM. Chapter 6 describes the BOWFOR methodology and its application to a ship journey from Europe to North America and return. Ship visual observations obtained during the trip are used to assimilate DOLPHIN, providing forecasts for the next points in the route. This application was based on synthetic data. The assimilation procedure used 4 ship observations per day during a 5-day assimilation period, and provided a 2-day forecast. This procedure was continuously performed along the journeys. This context is the most critical one may expect to have in practice, in the sense that there will usually be more information available, e.g. more visual observations, other ships' observations, or satellite or buoys observations, which may also be added to the assimilation scheme. Better results are expected if the number of observations of any source is increased, but what is of interest is that even in this critical setting the results were very good, specially on the trip from North America to Europe, when the ship was forerunning the wave conditions. Chapter 7 gathers achievements, conclusions and suggestions for further work.

Item Type: Thesis (MSc(R))
Qualification Level: Masters
Additional Information: Adviser: Carlos Soares Guedes
Keywords: Naval engineering
Date of Award: 1997
Depositing User: Enlighten Team
Unique ID: glathesis:1997-74750
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 27 Sep 2019 16:39
Last Modified: 27 Sep 2019 16:39
URI: https://theses.gla.ac.uk/id/eprint/74750

Actions (login required)

View Item View Item


Downloads per month over past year