01whole.pdf (3.58 MB)
Download file

A comparative analysis of remote sensing methods for burn-scar mapping in wetlands

Download (3.58 MB)
thesis
posted on 28.03.2022, 21:09 authored by Mackenzie Austin
This thesis examines the accuracy and reliability of using conventional remote sensing methods in wetland environments, specifically for detecting and analysing burn scars and post-fire recovery. To date, the methods have been applied in dryland environments post-fire but have yet to be tested comprehensively elsewhere, particularly in large floodplain wetland environments that also experience regular wildfires. These ecosystems possess unique physical characteristics that may pose challenges to orthodox remote sensing techniques, such as different vegetation composition, soil moisture prevalence, and post-fire vegetation regrowth rate. This thesis also aims to provide a preliminary insight into how the current methods could be modified or further developed to achieve the high levels of accuracy gained in measuring burn scars in dryland environments. It was found that spectral indices that measure green vegetation cover, particularly the GNDVI, the EVI2, the NBRI, and the SATVI, are more suitable for detecting burn. It was also found that supervised classifiers, were more accurate than unsupervised classifiers at determining burned areas. From the results, it was concluded that, by a measure of the assessed metrics, post-fire wetland ecosystems typically regenerate to pre-fire conditions within a year of the fire event occurring. However, different conditions can significantly decrease the time taken for regeneration, with some indices returning to pre-fire values as early as 100 days post-fire.

History

Table of Contents

1. Preface -- 2. Journal paper -- 3. Synthesis -- References -- Appendix.

Notes

Bibliography: pages 51-58 Empirical thesis.

Awarding Institution

Macquarie University

Degree Type

Thesis MRes

Degree

MRes, Macquarie University, Faculty of Science and Engineering, Department of Environmental Sciences

Department, Centre or School

Department of Environmental Sciences

Year of Award

2017

Principal Supervisor

Michael Chang

Additional Supervisor 1

Kerrie Timkins

Additional Supervisor 2

Timothy John Ralph

Rights

Copyright Mackenzie Austin 2017. Copyright disclaimer: http://mq.edu.au/library/copyright

Language

English

Extent

1 online resource (viii, 66 pages) colour illustrations, colour maps

Former Identifiers

mq:70636 http://hdl.handle.net/1959.14/1266221