In questa tesi vengono analizzate le serie temporali relative ad ampiezze di onde oceaniche registrate a Currituck Sound, in North Carolina, nel corso di un esperimento condotto dai ricercatori Donald T. Resio e Chuck Long del centro FRF di Vicksburg, in Missisipi. I dati analizzati furono misurati durante una tempesta,e testimoniano la presenza di onde anomale. Eseguendo una statistica lineare, si dimostra che le onde anomale, in oceano, non possono essere predette basandosi su una distribuzione gaussiana di probabilità, come, al contrario, è d'uso fare in oceanografia. Successivamente, attraverso l'utilizzo della trasformata di Fourier (FFT), l'informazione nel dominio del tempo viene convertita in informazioni di ampiezza e fase nel dominio della frequenza. Sostituendo, a questo punto, nel dominio della frequenza, le fasi originali delle onde, con fasi random generate numericamente, si scopre che, proprio in seguito a tale sostituzione, la statistica delle onde oceaniche segue ora una distribuzione gaussiana di probabilità. Per questa ragione, è ragionevole ipotizzare che le fasi siano legate fra loro, e che sia proprio questo legame la causa della presenza di onde anomale nelle serie temporali. Inoltre, dagli spettri di Fourier, risulta esserci una netta separazione in frequenza, in corrispondenza di una frequenza di taglio pari a ω = 0.3 Hz. Questa separazione mette in risalto il fatto che la generazione di onde anomale può essere scatenata da due tipi diversi di instabilità: una legata alla soluzione di onda solitaria dell'equazione di Korteweg-De Vries, e valida per frequenze inferiori alla frequenza di taglio, l'altra legata all'evoluzione di pacchetti d'onda instabili in onde anomale, come descritta dall'equazione di Nonlinear Shroedinger, e valida per frequenze superiori alla frequenza di taglio. Queste soluzioni instabili, per le equazioni di Kortewg-De Vries e Nonlinear Shroedinger, richiedono l'esistenza di un certo grado di correlazione fra le fasi. Per comprendere se uno dei due tipi di instabilità, a bassa o ad alta frequenza, abbia il maggior contributo nella generazione di onde anomale, vengono condotte delle prove, applicando, in maniera combinata, filtri passa basso e passa alto, e inserimento di fasi random. Da questo lavoro di ricerca si può concludere che le onde oceaniche non seguono una distribuzione gaussiana, e che ciò è provocato da una connessione fra le fasi delle onde stesse. Inoltre, sembrano essere le onde ad alta frequenza, o meglio le instabilità che hanno origine ad alta frequenza, ad avere il maggior impatto sulla formazione di onde anomale.
This thesis is an analysis of observations of ocean waves obtained in North Carolina in the United States of America. The experiment was conducted by the FieldResearch Facility in Currituck Sound, North Carolina by Donald T. Resio and Chuck Long of the U. S. Army Corp of Engineers. The wave data analyzed here were collected during a large Atlantic ocean storm and are unique in the precision and specific location of the instrumentation which was placed in a water depth of 2.2 meters. A linear statistical analysis has been performed and the data are found to be highly non-Gaussian, a result which suggests the presence of large nonlinear effects in the data. The fast Fourier transform (FFT) is used to convert the time-domain data into Fourier amplitudes and phases in the frequency domain. The method of ¿surrogates¿ is then used to test the data for nonlinear effects. In this approach the Fourier phases are replaced by numerically generated random phases. By inverting the surrogate Fourier spectrum back to the time domain it has been found that the statistics of the wave amplitudes are Gaussian. For this reason, it seems reasonable to assume that the phases of the original data are nonlinearly correlated and the resulting phase locking is the main cause for nonlinear effects in the recorded time series. Furthermore, it has been found that a well-defined separation appears in the frequency domain between ¿low frequency¿ shallow-water behavior and ¿high frequency¿ deep-water behavior. This is because there is a well-defined minimum in the spectrum near ω = 0.3 Hz, a result that depends primarily on the depth selected for the instrumentation. The physical behavior to the left of ω is dominated by low-frequency behavior described to leading order by the Korteweg-De Vries equation, which has soliton solutions. To the right of ω, the dynamics are described to leading order by the nonlinear Schroedinger equation (the Fourier spectrum is quite peaked), which has ¿breather¿ type solutions, often interpreted as rogue waves. We have found, using traditional filtering methods, that these interpretations are essentially correct. Indeed we find solitons in the low-frequency part of the spectrum and ¿unstable, rogue wave type¿ packets in the high frequency part of the spectrum. From this research effort, we can conclude that ocean waves do not follow a normal probability distribution law, and this is caused by phase locking in the same waves. Moreover, it appears that it is the high frequencies waves, or better instabilities that grows at high frequencies, who have the largest impact on the rogue waves formation. These results, seemingly quite surprising, appears to be rather robust and points the way to further research in the future to better understand the nonlinear dynamics of solitons with unstable wave packets, the most important coherent structures in ocean surface waves.
Analisi dati di onde oceaniche registrate durante l'esperimento condotto a Currituck Sound, North Carolina
BELLOMO REPETTO, KATINKA
2009/2010
Abstract
This thesis is an analysis of observations of ocean waves obtained in North Carolina in the United States of America. The experiment was conducted by the FieldResearch Facility in Currituck Sound, North Carolina by Donald T. Resio and Chuck Long of the U. S. Army Corp of Engineers. The wave data analyzed here were collected during a large Atlantic ocean storm and are unique in the precision and specific location of the instrumentation which was placed in a water depth of 2.2 meters. A linear statistical analysis has been performed and the data are found to be highly non-Gaussian, a result which suggests the presence of large nonlinear effects in the data. The fast Fourier transform (FFT) is used to convert the time-domain data into Fourier amplitudes and phases in the frequency domain. The method of ¿surrogates¿ is then used to test the data for nonlinear effects. In this approach the Fourier phases are replaced by numerically generated random phases. By inverting the surrogate Fourier spectrum back to the time domain it has been found that the statistics of the wave amplitudes are Gaussian. For this reason, it seems reasonable to assume that the phases of the original data are nonlinearly correlated and the resulting phase locking is the main cause for nonlinear effects in the recorded time series. Furthermore, it has been found that a well-defined separation appears in the frequency domain between ¿low frequency¿ shallow-water behavior and ¿high frequency¿ deep-water behavior. This is because there is a well-defined minimum in the spectrum near ω = 0.3 Hz, a result that depends primarily on the depth selected for the instrumentation. The physical behavior to the left of ω is dominated by low-frequency behavior described to leading order by the Korteweg-De Vries equation, which has soliton solutions. To the right of ω, the dynamics are described to leading order by the nonlinear Schroedinger equation (the Fourier spectrum is quite peaked), which has ¿breather¿ type solutions, often interpreted as rogue waves. We have found, using traditional filtering methods, that these interpretations are essentially correct. Indeed we find solitons in the low-frequency part of the spectrum and ¿unstable, rogue wave type¿ packets in the high frequency part of the spectrum. From this research effort, we can conclude that ocean waves do not follow a normal probability distribution law, and this is caused by phase locking in the same waves. Moreover, it appears that it is the high frequencies waves, or better instabilities that grows at high frequencies, who have the largest impact on the rogue waves formation. These results, seemingly quite surprising, appears to be rather robust and points the way to further research in the future to better understand the nonlinear dynamics of solitons with unstable wave packets, the most important coherent structures in ocean surface waves.File | Dimensione | Formato | |
---|---|---|---|
272509_tesi-bellomo-2010.pdf
non disponibili
Tipologia:
Altro materiale allegato
Dimensione
2.99 MB
Formato
Adobe PDF
|
2.99 MB | Adobe PDF |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14240/15766