Spatial analysis of forest fuels for wildland fire behaviour in south-eastern Australia

Abstract

Fire influences species composition and vegetation structure in forests and woodlands. Fire also affects carbon stores, influences soil biochemistry, and alters hydrology (e.g. Bowman, Williamson, Price, et al. 2021). Fire is therefore inextricably linked to forest functions. In south-east Australia, fires are also a safety concern in a growing wildland-urban interface (Price and Bradstock 2013; Price et al. 2015; Ghaderi et al. 2021). Fires are expected to become larger, more severe, and more frequent under climate change in south-east Australia (Van Oldenborgh et al. 2021; Abram et al. 2021; Canadell et al. 2021). Understanding the drivers of forest dynamics and forest fires in the region is of critical importance to better predicting the degree of resistance and resilience of forests and woodlands to changing disturbance regimes. In this thesis, I quantified the relative importance of environmental and pyro-geographic constraints on eucalypt bark type distributions, using machine learning and existing species occurrence records and environmental data. I found that temperature, water availability, and fire history influenced the distributions of eucalypt bark types. I demonstrated support for theories of trade-offs in investment between the water use efficiency of stem photosynthesis provided by smooth bark and the fire protection provided by rough bark. Next, I examined the influences of near-term drought, time-since-fire and fire severity on forest structure. For this analysis, I used Fire Extent and Severity Mapping (FESM; Gibson et al. 2020) data products to represent fire severity, and Global Ecosystem Dynamics Investigation (GEDI; Tang and Armston 2019) data products to represent forest structure. I found that recovery of forest structure following fire was delayed by post-fire near-term drought. Lastly, I used results from these analyses to quantify the influences of landscape, fuel, and weather variables in determining fire behaviour during the Black Summer fires of 2019 – 20, taking advantage of novel data products created for the New South Wales Bushfire Inquiry (Final Report of the NSW Bushfire Inquiry 2020). I found that topographic complexity was the strongest predictor of fire progression rates during the Black Summer fires, following by understorey cover percentage and live fuel moisture content. The work completed for this thesis provides new insights into eucalypt ecology, forest structure, and fire behaviour under climate change. The information and data products produced in this thesis can be readily applied to improve operational fire management across New South Wales.

Date of Award	2024
Original language	English
Awarding Institution	Western Sydney University
Supervisor	Matthias Boer (Supervisor), Rachael Nolan (Supervisor), Mark Tjoelker (Supervisor), Michael Aspinwall (Supervisor) & Kim Calders (Supervisor)