Thesis
The uses of drone mapping to reveal aquatic environmental change
Southern Cross University
Doctor of Philosophy (PhD), Southern Cross University
2022
Metrics
110 File views/ downloads
135 Record Views
Abstract
Anthropogenic pressures have affected every natural ecosystem on the planet. As these pressures compound, researchers need effective tools to monitor environments and catalogue change, to best inform management actions and conservation strategies. Remote sensing techniques are important for environmental monitoring, but satellites and crewed aircraft are limited by temporal resolution, cost and environmental factors. Drone-based orthomosaic imagery (maps) have the potential to provide high-resolution imagery from low altitudes and user-controlled temporal sampling at reasonable cost and scale. Novel drone technology must be tested to develop best-practice mapping and image analysis techniques. This is particularly important in environments with added remote sensing challenges, such as aquatic habitats.
My thesis explores and addresses the uses and limitations of RGB (red, green, blue) and multispectral drone-based orthomosaic imagery across coastal, estuarine and freshwater aquatic environments. I developed an automated method of classifying sun glint in aquatic imagery to resolve image contamination (Chapter 2). This method used object-based image analysis and deep learning techniques, trained on RGB drone-based imagery. I then expanded these methods to a coastal shallow-water reef environment, creating an automated system to classify bleached and unbleached corals for drone-based RGB imagery (Chapter 3). To test the versatility of drone imagery, Chapter 4 was conducted in a dynamic estuarine environment. Here, I tested the potential for multispectral imagery to predict a range of water quality parameters, and assessed if a model could be developed to capture temporal fluctuations in water quality. Drone imagery predicted spatial variation in water quality, however a temporally robust model of water quality could not be produced. Finally, I tested the potential applications of RGB drone-based orthomosaic imagery around an alpine freshwater ecosystem impacted by feral horses (Chapter 5). I mapped and classified a range of horse damage factors relating to riparian vegetation and stream health across a gradient of horse presence. Drones were found to be an effective tool to quantify feral horse impact. A clear correlation between drone-derived horse impact classifications and horse presence was identified.
Overall, drones are a useful remote sensing tool for aquatic and riparian environments, providing low cost, high-resolution imagery. I determined some limitations of drone imagery for temporal analysis, as well as further improvements to mapping software and the development of streamlined imagery analysis options. My thesis offers some image analysis options and demonstrates that drones are an effective tool for spatial ecology in a range of environments.
Details
- Title
- The uses of drone mapping to reveal aquatic environmental change
- Creators
- Anna B Giles
- Contributors
- Brendan Kelaher (Supervisor) - Southern Cross UniversityIsaac R Santos (Supervisor) - Southern Cross University
- Awarding Institution
- Southern Cross University; Doctor of Philosophy (PhD)
- Theses
- Doctor of Philosophy (PhD), Southern Cross University
- Publisher
- Southern Cross University
- Number of pages
- xviii, 154
- Identifiers
- 991013080613802368
- Copyright
- © AB Giles 2022
- Academic Unit
- Faculty of Science and Engineering
- Resource Type
- Thesis