dc.contributor.advisor |
Ford, M |
en |
dc.contributor.author |
Holdaway, Andrew |
en |
dc.date.accessioned |
2018-10-09T02:01:47Z |
en |
dc.date.issued |
2018 |
en |
dc.identifier.uri |
http://hdl.handle.net/2292/39746 |
en |
dc.description |
Available to authenticated members of The University of Auckland. |
en |
dc.description.abstract |
Atoll islands are deposits of reef-derived sediment located on atoll rims throughout the tropical and subtropical oceans, representing the only inhabitable land area in countries such as the Maldives and Tuvalu. Low lying and of limited extent, atoll islands are widely perceived to be highly vulnerable to climate change and associated sea level rise. Predictions of inundation and chronic erosion have led to concern for the future habitability of these features, and the potential for residents to become the world’s earliest climate change refugees. To date, attempts to quantify atoll island shoreline change over time have been based on the manual interpretation of recent very high-resolution satellite imagery and historic aerial photographs. Due to poor availability and the costs of imagery purchase and digitisation, existing studies have seldom taken place beyond the atoll scale, with sparse observations through time. This study develops a large-scale remote sensing based approach which avoids these limitations. Using the cloud-based remote sensing platform Google Earth Engine (GEE), the change in land area of 223 atolls across the Pacific Ocean, Indian Ocean and the South China Sea was measured. All available Landsat-7 and Landsat-8 TOA imagery was cloud-masked and combined to form a time-series of composite images from 2000 to 2017. Where the final composite was of low quality or no land was detected, atolls were eliminated, leaving 168 in the final analysis. Of the available classifiers, Support Vector Machine (SVM) provided the best overall performance. Composite classification was achieved via SVM and a simple scheme of covers such as vegetated or urban land, submerged reef etc. A reference dataset was generated via manual interpretation of high-resolution imagery to perform accuracy assessment. The overall accuracy of the classification varied between 77 and 96%. Results of the classification were then aggregated to form a binary classification of land and non-land covers. Results reveal the total net landmass of the atolls examined has increased from 901.4 km2 to 960.7 km2 between 2000 and 2017. 91% of the additional atoll island landmass added globally has been because of anthropogenic island-building and reclamation activities within the Maldives (37.50 km2 ) and in the South China Sea (16.55 km2 ), with no evidence of chronic loss occurring |
en |
dc.publisher |
ResearchSpace@Auckland |
en |
dc.relation.ispartof |
Masters Thesis - University of Auckland |
en |
dc.relation.isreferencedby |
UoA99265112012702091 |
en |
dc.rights |
Items in ResearchSpace are protected by copyright, with all rights reserved, unless otherwise indicated. Previously published items are made available in accordance with the copyright policy of the publisher. |
en |
dc.rights |
Restricted Item. Available to authenticated members of The University of Auckland. |
en |
dc.rights.uri |
https://researchspace.auckland.ac.nz/docs/uoa-docs/rights.htm |
en |
dc.rights.uri |
http://creativecommons.org/licenses/by-nc-sa/3.0/nz/ |
en |
dc.title |
A cloud-based remote sensing approach to measuring atoll island change at a global scale |
en |
dc.type |
Thesis |
en |
thesis.degree.discipline |
Geography, |
en |
thesis.degree.grantor |
The University of Auckland |
en |
thesis.degree.level |
Masters |
en |
dc.rights.holder |
Copyright: The author |
en |
pubs.elements-id |
754583 |
en |
pubs.org-id |
Science |
en |
pubs.org-id |
School of Environment |
en |
pubs.record-created-at-source-date |
2018-10-09 |
en |
dc.identifier.wikidata |
Q112936664 |
|