In the past decades, the use of remote sensing, and in particular the application of satellite images, has revolutionized the study of fluvial morphological processes. This has made it possible to study riverbed processes on a large scale, whether geographical or temporal. However, little is still known about the dynamics of sediment bars, which are the subject of this thesis. This research is based on the use of Google Earth Engine (GEE), a cloud platform that enables the processing, visualization, and experimentation with satellite images from both Landsat and Sentinel satellites. The study draws inspiration from the paper by Boothroyd et al. (2020), Applications of Google Earth Engine in fluvial geomorphology for detecting river channel change, to specifically analyse the temporal evolution of sediment bars in two reaches of the Po River starting from the 1990s, using images from the Landsat collection. Thanks to the data from the Landsat collection, it was possible to create a time series covering over 30 years. The work involved adapting the base code to the specific case study, extracting satellite images from GEE, and analysing the resulting data in QGIS and with Python scripts. The information acquired on sediment bars was then compared with hydrological variables to identify possible factors that influenced geomorphological changes. The results show an areal decrease of bars in both study areas, with a deceleration of this process that has been going on since 2017. Furthermore, the results derived from Landsat images were compared with those obtained from the Sentinel collection, particularly from the Sentinel-2 satellite, to assess the effect of the sensors' different spatial resolutions. This research demonstrates the growing effectiveness of remote sensing in fluvial geomorphological analysis, highlighting the advantages of a free platform like GEE, which offers a vast amount of data accessible to anyone. This tool allows for the observation of river processes' evolution and, in the specific case study, sheds light on the dynamics of the Po River, particularly the trajectories of its sediment bars.
Nell’ultimo decennio l’utilizzo del remote-sensing ed in particolare, l’impiego delle immagini satellitari, ha rivoluzionato lo studio dei processi morfologici fluviali. Questo ha reso possibile lo studio dei processi dell’alveo su vasta scala, sia essa geografica o temporale. Tuttavia, poco ancora si sa sulle dinamiche delle barre di sedimento, oggetto di questa tesi. Questa ricerca si basa sull'uso di Google Earth Engine (GEE), una piattaforma cloud che consente l'elaborazione, la visualizzazione e la sperimentazione con immagini satellitari, provenienti sia dai satelliti Landsat che Sentinel. Lo studio prende ispirazione dall'articolo di Boothroyd et al. (2020), Applications of Google Earth Engine in fluvial geomorphology for detecting river channel change, per analizzare nello specifico l'evoluzione nel tempo delle barre di sedimento in tre tratti del fiume Po a partire dagli anni '90, utilizzando le immagini della collezione Landsat. Grazie ai dati della collection Landsat è stato possibile ricavare una serie temporale che copre oltre 30 anni. Il lavoro ha previsto l'adattamento del codice di base al caso studio in questione, l'estrazione delle immagini satellitari da GEE e l'analisi dei dati risultanti. Le informazioni acquisite relative alle barre di sedimento sono state successivamente confrontate con variabili idrologiche per identificare i possibili fattori che hanno influenzato le modificazioni geomorfologiche. Inoltre, i risultati derivanti dalle immagini Landsat sono stati confrontati con quelli ottenuti dalla collezione Sentinel, in particolare dal satellite Sentinel-2, per valutare le differenze dovute alla diversa risoluzione spaziale dei sensori. Questa ricerca dimostra l'efficacia crescente del remote-sensing nell'analisi geomorfologica fluviale, evidenziando l'utilità di una piattaforma gratuita come GEE, che offre una vasta quantità di dati accessibili a chiunque. Tale strumento permette di osservare l'evoluzione dei processi fluviali e, nel caso studio specifico, di fare luce sulle dinamiche del fiume Po, in particolare, sulle traiettorie delle sue barre di sedimento.
Evoluzione delle barre di sedimenti fluviali tramite un classificatore semi-automatico utilizzando immagini satellitari
MESTRINER, LUCA
2023/2024
Abstract
In the past decades, the use of remote sensing, and in particular the application of satellite images, has revolutionized the study of fluvial morphological processes. This has made it possible to study riverbed processes on a large scale, whether geographical or temporal. However, little is still known about the dynamics of sediment bars, which are the subject of this thesis. This research is based on the use of Google Earth Engine (GEE), a cloud platform that enables the processing, visualization, and experimentation with satellite images from both Landsat and Sentinel satellites. The study draws inspiration from the paper by Boothroyd et al. (2020), Applications of Google Earth Engine in fluvial geomorphology for detecting river channel change, to specifically analyse the temporal evolution of sediment bars in two reaches of the Po River starting from the 1990s, using images from the Landsat collection. Thanks to the data from the Landsat collection, it was possible to create a time series covering over 30 years. The work involved adapting the base code to the specific case study, extracting satellite images from GEE, and analysing the resulting data in QGIS and with Python scripts. The information acquired on sediment bars was then compared with hydrological variables to identify possible factors that influenced geomorphological changes. The results show an areal decrease of bars in both study areas, with a deceleration of this process that has been going on since 2017. Furthermore, the results derived from Landsat images were compared with those obtained from the Sentinel collection, particularly from the Sentinel-2 satellite, to assess the effect of the sensors' different spatial resolutions. This research demonstrates the growing effectiveness of remote sensing in fluvial geomorphological analysis, highlighting the advantages of a free platform like GEE, which offers a vast amount of data accessible to anyone. This tool allows for the observation of river processes' evolution and, in the specific case study, sheds light on the dynamics of the Po River, particularly the trajectories of its sediment bars.File | Dimensione | Formato | |
---|---|---|---|
Mestriner_Luca.pdf
accesso aperto
Dimensione
4.83 MB
Formato
Adobe PDF
|
4.83 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/78232