Unit:
Department of PhysicsLibrary of the School of Science
Author:
Metheniti Vasiliki
Dissertation committee:
Σαράντης Σοφιανός, Αναπληρωτής Καθηγητής, Τμήμα Φυσικής, ΕΚΠΑ,
Νικόλαος Καμπάνης, Διευθυντής Ερευνών, Ινστιτούτο Υπολογιστικών Μαθηματικών, Ίδρυμα Τεχνολογίας και Έρευνας (ΙΤΕ)
Αριστομένης Καραγεώργης, Διευθυντής Ερευνών, Ινστιτούτο Ωκεανογραφίας, Ελληνικό Κέντρο Θαλασσίων Ερευνών (ΕΛΚΕΘΕ)
Παναγιώτης Δρακόπουλος, Καθηγητής, Τμήμα Βιοϊατρικών Επιστημών, Πανεπιστήμιο Δυτικής Αττικής
'Ελενα Φλόκα, Καθηγήτρια, Τμήμα Φυσικής, ΕΚΠΑ,
Ελισσάβετ Μποσιώλη, Επίκουρη Καθηγήτρια, Τμήμα Φυσικής, ΕΚΠΑ,
Γεωργία Σωτηροπούλου, Επίκουρη Καθηγήτρια, Τμήμα Φυσικής, ΕΚΠΑ
Original Title:
The impact of the optical characteristics of the sea surface layer on the ocean dynamics (combining observations and models in the Aegean Sea)
Translated title:
The impact of the optical characteristics of the sea surface layer on the ocean dynamics (combining observations and models in the Aegean Sea)
Summary:
The Aegean Sea’s turbidity field is influenced by the qualitative and quantitative distribution of suspended particles and dissolved organic matter in the basin, along with the hydrodynamic characteristics of the area. Scope of this dissertation is to study the effects of turbidity in the surface dynamics and characteristics of the Aegean Sea. In order to investigate the responsible mechanisms and answer the scientific questions raised during the preparation of the thesis, observational data, machine learning algorithms and ocean circulation numerical models were used.
As part of the study of the effect of turbidity on the surface features of the Aegean, a turbidity dataset was created for applications in ocean simulation models. For the derivation of the turbidity gridded field, a machine learning algorithm was developed and applied, utilizing in-situ optical measurements in the Aegean Sea, provided by HCMR (Hellenic Centre for Marine Research) for the period 1991-2019. Additionally, ocean circulation twin experiment simulations were performed for the period 1997-2001, using a high-resolution 1/36ο ocean model. The turbidity field based on the in-situ measurements and the corresponding satellite product were used as input for the solar radiation penetration parameterization. The analysis focused on the average value of the last year of the experiments.
The final turbidity product of the machine learning algorithm is consistent with the surface distribution characteristics of the in-situ optical measurements and with the available literature. It also highlights the contribution of dissolved organic matter, which often affects satellite measurements and derived turbidity products in such regions. In particular, the highest turbidity values appear in the northern Aegean and are mainly due to the local primary production and the presence of suspended particles and organic dissolved matter, originating from river outflows and the Black Sea water mass surface inflow in the east. The relationship of turbidity with the Aegean’s surface circulation and the influence of meso-scale circulation features are also shown. Compared to the corresponding satellite product, while the latter is generally less turbid, the field that is based on the in-situ measurements shows a sharper gradient. This field appears strengthened in the northern Aegean and better captures the regional meso-scale circulation especially in the northeast, near the Samothraki anticyclone, which is attributed to the local excess in dissolved organic matter.
The twin-experiment simulations highlighted the need for a more realistic approach of turbidity in ocean numerical simulations. Turbidity was shown to affect the sea surface temperature and salinity, as well as the surface circulation features of the basin, while two mechanisms were identified to account for these changes. The first concerns the direct connection of turbidity and sea surface temperature, which occurs mainly in coastal areas, and on the continental shelf of the North Aegean. The second mechanism concerns circulation feedback, which is due to the indirect influence of turbidity on parameters of the air-sea interaction layer. A result of this mechanism is the increased anticyclonic activity of the northeastern Aegean in the experiment that used the turbidity dataset derived from in-situ observations. The enhanced intensity anticyclonic features, like the Samothraki anticyclone, result in prolonged circulation of the Black Sea water mass in the northern basin, indicating its relative decoupling from the southern basin, as they are separated from the Cyclades Plateau.
The results of this dissertation can contribute to the study of Aegean’s dynamics, as well as the dynamics of the broader Mediterranean basin, and of regions with a similar optical status. The methodologies developed, i.e., the machine learning algorithm and the turbidity derivation methodology, can also be used in the design of generating turbidity fields and other optical and biogeochemical parameters of the ocean.
Main subject category:
Science
Keywords:
Aegean Sea, turbidity, ocean color, numerical model, NEMO, machine learning
Number of references:
163
File:
File access is restricted until 2025-01-04.
Doctoral_Dissertation_Metheniti_V.pdf
3 MB
File access is restricted until 2025-01-04.
