The application of object-based image analysis to geomorphological seabed mapping

Simple item page
dc.contributor.advisorWheeler, Andrew
dc.contributor.advisorLim, Aaron
dc.contributor.authorSummers, Gerarden
dc.contributor.funderEuropean Commission
dc.contributor.funderInterreg
dc.contributor.funderIrish Research Council
dc.contributor.funderGeological Survey of Ireland
dc.date.accessioned2023-05-24T15:37:14Z
dc.date.available2023-05-24T15:37:14Z
dc.date.issued2023-02
dc.date.submitted2023-02
dc.description.abstractIncreasing anthropogenic pressures on marine ecosystems due to the reliance on marine resources and the intense development of the marine realm within the last 10 years has threatened the effective functioning of many unique and fragile marine habitats. These environmental stresses warrant effective monitoring and management practices to ensure the preservation of good environmental status. In situ monitoring of marine environmental processes, such as current flow analysis through Acoustic Doppler Current Profilers, provide data with a high temporal resolution at distinct points on the seafloor. However, extensive spatial coverage of seafloor environmental requires a more efficient strategy to quantify these processes. Seabed mapping has long been established as an essential implement in the effective administration of marine ecosystems. Furthermore, in recognition of the significance of seabed mapping in the successful governance of the marine realm, several international seabed mapping initiatives and national seabed mapping programmes have been established with the goal to achieve complete mapping coverage of the seafloor by 2030. Such a significant volume of data evokes the necessity for an objective and repeatable approach to extract meaningful information from the seabed. Geomorphological seafloor features, including current induced seabed sedimentary bedforms (SSBs), are important indicators of habitat, and are readily apparent in seabed mapping data. Moreover, SSBs are common to many marine habitats and spatial scales and are the physical expression of seafloor hydrodynamics, thus these features are appropriate for a standardised approach designed to ascertain objective information on seabed hydrodynamics. This thesis develops a scale robust object-based image analysis (OBIA) approach that is created to classify SSBs as depicted in multibeam echosounder (MBES) bathymetry and derive hydrodynamic information from their morphometrics. This OBIA approach was applied to SSBs occurring in two spatial resolutions of MBES data. Here, four machine learning classifiers support vector machines, two multi-layer perceptrons, and voting ensemble were assessed on their ability to classify SSBs in these two resolutions of data. The results show that the voting ensemble classifier provided the most accurate results for both datasets. The OBIA framework was applied to SSBs depicted in MBES data acquired in a cold-water coral (CWC) habitat in the “downslope Moira Mounds” in the Porcupine Seabight. The SSB attributes of wave height and wavelength were derived from the SSBs classified in these data were used as an input to a multiple linear regression that predicted the seabed current velocity. These predictions illustrated the variable influence of topographic steering occurring at regional-, local-, and micro-spatial scales on regional hydrodynamics. This workflow presented the first estimation of current flow velocity and direction from SSBs in MBES data. Finally, this OBIA approach was used to assess SSBs occurring in multiple resolutions of data within the same region, altering the resolution of observation to evaluate the effect of spatial resolution on the temporal resolution of seabed current hydrodynamics. Moreover, this study determined that the coarse spatial resolution MBES data prevented the assessment of short-term variations in seabed benthic habitat hydrodynamics.
dc.description.statusNot peer revieweden
dc.description.versionAccepted Versionen
dc.format.mimetypeapplication/pdfen
dc.identifier.citationSummers, G. 2023. The application of object-based image analysis to geomorphological seabed mapping. PhD Thesis, University College Cork.
dc.identifier.endpage175
dc.identifier.urihttps://hdl.handle.net/10468/14509
dc.language.isoenen
dc.publisherUniversity College Corken
dc.rights© 2023, Gerard Summers.
dc.rights.urihttps://creativecommons.org/licenses/by-nc-nd/4.0/
dc.subjectObject-based image analysis
dc.subjectGeomorphology
dc.subjectSeabed sedimentary bedform
dc.titleThe application of object-based image analysis to geomorphological seabed mapping
dc.typeDoctoral thesisen
dc.type.qualificationlevelDoctoralen
dc.type.qualificationnamePhD - Doctor of Philosophyen
Files
Original bundle
Now showing 1 - 2 of 2
Thumbnail Image
Name:
SummersG_PhD2023.pdf
Size:
8.97 MB
Format:
Adobe Portable Document Format
Description:
Full Text E-thesis
Download
Thumbnail Image
Name:
Submission for examination form.pdf
Size:
418.46 KB
Format:
Adobe Portable Document Format
License bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
license.txt
Size:
5.2 KB
Format:
Item-specific license agreed upon to submission
Description:
Download
Collections
Research Theses Submission
Biological, Earth and Environmental Sciences - Doctoral Theses
Centre for Marine and Renewable Energy (MaREI) - Doctoral Theses
College of Science, Engineering and Food Science - Doctoral Theses