| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 28.11 MB | Adobe PDF |
Authors
Abstract(s)
A presente Dissertação de Mestrado visa a produção de uma ferramenta
informática, em Matlab, para análise e previsão de marés e correntes de maré, a
qual foi designada SeaTideWave. Pretende-se a partir de três métodos matemáticos
- séries de Fourier e duas variantes de mínimos quadrados - Boon e Kiley (1978)
e Boon (2004) – calcular os coeficientes de Fourier que ajustem o polinómio trigonométrico às observações e obter as constantes harmónicas, as quais serão alvo de
uma avaliação de qualidade. Procura-se encontrar uma metodologia para detetar e
avaliar a ocorrência de eventos de maré meteorológica, como storm surges.
Foi criado um sistema de programas que envolvem: a análise harmónica
e previsão com recurso àqueles métodos, a partir de registos temporais entre 29 a
369 dias; controlo de qualidade das constantes harmónicas conseguidas, de vários
portos ou locais, envolvendo o tratamento estatístico dos resíduos temporais entre
as observações e as correspondentes previsões, complementado por análise espectral;
e deteção e avaliação de marés meteorológicas, a partir de processos de filtragem e
espectros cruzados, de registos de Newport, EUA, que incluem registos de pressão
atmosférica e ventos.
Os resultados obtidos revelam a boa qualidade dos modelos preditivos para
as alturas de maré, em portos costeiros onde os efeitos da propagação em pequenas
profundidades não sejam significativos, e razoável qualidade no caso de correntes.
A partir da metodologia inerente à deteção e avaliação de storm surges constata-se
o efeito de «barómetro invertido», entre os episódios de maré meteorológica e da
pressão atmosférica, bem como a consistência da teoria de Ekman para explicar
esses episódios.
Destaca-se vários aspetos inovadores, associados à ferramenta SeaTideWave,
nomeadamente a implementação da análise por mínimos quadrados e a capacidade
de previsão de correntes de maré, demonstrada pela «tabela de correntes» para 2022
e também na forma de «elementos de correntes de maré», para um local próximo da
Trafaria, no estuário do Tejo
This Master’s Thesis aims to produce a computer tool, in Matlab, for the analysis and prediction of tides and tidal currents, which was named SeaTideWave. From three mathematical methods - Fourier series and two least-squares variants - Boon and Kiley (1978) and Boon (2004) - calculates the Fourier coefficients, that adjusts the trigonometric polynomial to the observations and obtains the harmonic constants, which will be subject to a quality assessment. It seeks to find a methodology to detect and assess the occurrence of meteorological tidal events, such as storm surges. A system of programs was created that involves: harmonic analysis and prediction using those methods, based on time series between 29 and 369 days; quality control of the harmonic constants achieved, from various ports or locations, involving the statistical treatment of residues, between the observations and the corresponding predictions, complemented by residues spectral analysis; and detection and evaluation of meteorological tides, using filtering processes and cross spectra, using Newport, USA, atmospheric pressure and wind data. The obtained results reveal the quality of the predictive models for the tide height in coastal ports where the propagation effects in shallow waters are not significant and reasonable for the currents. Based on the methodology inherent to the detection and evaluation of storm surges, the «inverted barometer effect» can be seen between the meteorological tide and atmospheric pressure episodes, as well as the consistency of Ekman’s theory to explain these episodes. There are several innovative aspects associated with the SeaTideWave tool, namely the implementation of the least-squares analysis and the ability to predict tidal currents, demonstrated by the «tidal current table» for 2022 and also in the form of «tidal current elements», for a place close to Trafaria, in the Tagus estuary.
This Master’s Thesis aims to produce a computer tool, in Matlab, for the analysis and prediction of tides and tidal currents, which was named SeaTideWave. From three mathematical methods - Fourier series and two least-squares variants - Boon and Kiley (1978) and Boon (2004) - calculates the Fourier coefficients, that adjusts the trigonometric polynomial to the observations and obtains the harmonic constants, which will be subject to a quality assessment. It seeks to find a methodology to detect and assess the occurrence of meteorological tidal events, such as storm surges. A system of programs was created that involves: harmonic analysis and prediction using those methods, based on time series between 29 and 369 days; quality control of the harmonic constants achieved, from various ports or locations, involving the statistical treatment of residues, between the observations and the corresponding predictions, complemented by residues spectral analysis; and detection and evaluation of meteorological tides, using filtering processes and cross spectra, using Newport, USA, atmospheric pressure and wind data. The obtained results reveal the quality of the predictive models for the tide height in coastal ports where the propagation effects in shallow waters are not significant and reasonable for the currents. Based on the methodology inherent to the detection and evaluation of storm surges, the «inverted barometer effect» can be seen between the meteorological tide and atmospheric pressure episodes, as well as the consistency of Ekman’s theory to explain these episodes. There are several innovative aspects associated with the SeaTideWave tool, namely the implementation of the least-squares analysis and the ability to predict tidal currents, demonstrated by the «tidal current table» for 2022 and also in the form of «tidal current elements», for a place close to Trafaria, in the Tagus estuary.
Description
Keywords
Marés Correntes Análise Harmónica Previsão Matlab