Braakmann-Folgmann, Anne Christina ORCID: https://orcid.org/0000-0003-4942-8545 (2023) Melting and fragmentation of Antarctic tabular icebergs tracked with satellite remote sensing and artificial intelligence. PhD thesis, University of Leeds.
Abstract
Icebergs release cold, fresh water and terrigenous nutrients as they drift and melt, influencing the local ocean properties and encouraging sea ice formation and biological production. To locate and quantify the freshwater flux from Antarctic icebergs, changes in their area and thickness have to be monitored along their trajectories. In this thesis, I developed novel methodology and datasets from satellite remote sensing observations to quantify the freshwater flux from giant Antarctic icebergs as accurately and efficiently as possible.
First, I investigated and improved the calculation of iceberg thickness from CryoSat-2 satellite altimetry observations. I found that employing consistently processed elevations is essential to detect changes in iceberg freeboard. Moreover, I developed a method to account for the evolution of the snow layer on icebergs during multi-annual drift and assessed its impact on iceberg thickness estimates. Combining these with measurements of iceberg area derived from satellite imagery, I estimate the volume loss (378 ± 57 km^3) and freshwater flux (106 ± 35 Gt) from the B30 iceberg over 6.5 years.
Next, I built on this methodology and applied it to the A68A iceberg, whose melting affected the ecosystem near South Georgia. I further improved the method by adding ICESat-2 data and automatically colocating altimetry tracks over the floating iceberg with a map of initial iceberg thickness. Overall, A68A released 802 ± 34 Gt of ice along its trajectory and 152 ± 61 Gt through basal melting near South Georgia.
Finally, I developed a deep neural network (based on U-net) to map the extent of giant icebergs in Sentinel-1 imagery. While each manual delineation takes several minutes, U-net reduces the time to 0.01 sec. Evaluating the performance compared to two standard segmentation techniques, I found that U-net achieves a higher F1 score (0.84 versus 0.62) and is more robust to sea ice, other icebergs and nearby coast.
Metadata
Supervisors: | Shepherd, Andrew and Hogg, David |
---|---|
Related URLs: |
|
Keywords: | tabular icebergs, satellite remote sensing, altimetry, CryoSat-2, ICESat-2, imagery, SAR, deep learning, artificial intelligence, neural network, U-net, mapping, Antarctica, Southern Ocean, freshwater flux, A68, B30 |
Awarding institution: | University of Leeds |
Academic Units: | The University of Leeds > Faculty of Environment (Leeds) > School of Earth and Environment (Leeds) |
Depositing User: | Ms Anne Braakmann-Folgmann |
Date Deposited: | 24 Jul 2023 13:17 |
Last Modified: | 17 Jan 2024 09:10 |
Open Archives Initiative ID (OAI ID): | oai:etheses.whiterose.ac.uk:32983 |
Download
Final eThesis - complete (pdf)
Filename: Braakmann-Folgmann_A_EarthEnvironment_PhD_2023.pdf
Licence:
This work is licensed under a Creative Commons Attribution NonCommercial ShareAlike 4.0 International License
Export
Statistics
You do not need to contact us to get a copy of this thesis. Please use the 'Download' link(s) above to get a copy.
You can contact us about this thesis. If you need to make a general enquiry, please see the Contact us page.