Search

Διερευνάται η επίδραση της κλιματικής αλλαγής στην εξέλιξη των μετεωρολογικών παλιρροιών στο Αιγαίο και το Ιόνιο Πέλαγος και την παράκτια ζώνη της Ελλάδας. Η ανάλυση καλύπτει την περίοδο 1951-2100 (κλιματικό σενάριο Α1Β για τoν 21ο αιώνα) και στηρίζεται σε προσομοιώσεις με ένα διδιάστατο υδροδυναμικό μοντέλο υψηλής χωρικής ανάλυση (GreCSSM). Η σύγκριση των προσομοιωμένων αποτελεσμάτων έναντι επί τόπου παρατηρήσεων με βάση στατιστικά μέτρα/δείκτες είναι ικανοποιητική. Παρουσιάζονται επίσης οι μελλοντικές τάσεις, η μεταβλητότητα και η συχνότητα εμφάνισης έντονων γεγονότων μετεωρολογικής παλίρροιας στις θαλάσσιες λεκάνες που βρέχουν την ελληνικές παράκτιες περιοχές. Δίνονται επίσης εκτιμήσεις για το σήμα της πιθανής κλιματικής αλλαγής μέχρι το 2100, μαζί τις μέσες, ετήσιες και υπερετήσιες ακραίες τιμές της θυελλογενούς ανύψωσης της θαλάσσιας στάθμης στο Αιγαίο και το Ιόνιο Πέλαγος., The impact of climate change on the evolution of storm surges in the Aegean and Ionian Seas (AIS) and the Greek coastal zone is studied. The analysis covers the 1951-2100 period (Α1Β climate scenario for the 21st century), and hinges on simulations with a two-dimensional hydrodynamic model of high spatial resolution (GreCSSM). The comparison of model results against in situ observations is satisfactory based on statistical measures and indices. Future trends, variability and occurrence frequency of severe storm surge events are also presented concerning the Greek maritime basins and coastal regions. Estimations are also provided for the probable climate change signal until 2100, along with averaged, intra- and inter-annual maxima of storm-induced sea level height of the Aegean and Ionian Seas.

Modelling Cohesive Sediment Dynamics in the Marine Environment

O mapeamento da vegetação, através da identificação do tipo e distribuição das comunidades e espécies vegetais, é crucial para analisar a cobertura vegetal e os padrões espaciais. A compreensão das variabilidades espaciais e temporais das plantas dunares em ligação com a morfodinâmica permite uma maior compreensão do dinamismo e evolução dos ambientes costeiros. Tal análise pode contribuir para o desenvolvimento de planos de gestão costeira que ajudam a implementar a biodiversidade costeira e estratégias de protecção. Esta dissertação apresenta uma abordagem para avaliar a utilização de imagens multiespectrais e explorar a variabilidade da vegetação dunar costeira com dados recolhidos à distância por um Veículo Aéreo Não Tripulado (UAV). Foram escolhidas quatro zonas de estudo diferentes na parte oriental da Península de Ancao, distribuídas alongshore, e cobrindo a backhore e a crista das dunas até à base do lee das dunas. Foram utilizados dados de campo e de UAV, em diferentes épocas, nomeadamente ao longo de um período de dois anos. Foi utilizada uma abordagem de classificação em duas etapas, baseada num índice de vegetação de diferença normalizada e num classificador de Floresta Aleatória. Os resultados mostram desempenhos de classificação de alta precisão ao condensar a cobertura do solo em menos classes e também em áreas menos densamente vegetativas. As classificações resultantes foram posteriormente processadas em termos de alterações transfronteiriças e alterações sazonais. Estas técnicas mostram um elevado potencial futuro para avaliar a vegetação das áreas de dunas costeiras e para apoiar a gestão costeira. The mapping of vegetation, by identifying the type and distribution of plant communities and species, is crucial for analysing vegetation coverage and spatial patterns. Understanding dune plant spatial and temporal variabilities in connection with morphodynamics gives further insight in dynamism and evolution of coastal environments. Such analysis can contribute to the development of coastal management plans that helps to implement coastal biodiversity and protection strategies. This dissertation presents an approach to assess the use of multispectral imagery and explore the variability of coastal dune vegetation with remotely sensed data collected by an Unmanned Aerial Vehicle (UAV). Four different study zones were chosen at the eastern part of the Ancao Peninsula, distributed alongshore, and covering the backshore and the dune crest until the base of the dune lee. Field and UAV data were used, in different seasons namely over an extend of two years. A two-step classification approach, based on a normalized difference vegetation index and Random Forest classifier, was used. The Results show high accuracy classification performances when condensing the groundcover into fewer classes and also in less densely vegetated areas. Resulting classifications were further processed in terms of cross-shore changes and seasonal changes. These technics show a high future potential to assess the vegetation of coastal dune areas and to support coastal management.