Logo image
Pyrogenic carbon emissions from the 2019 Yanchep wildfire
Thesis   Open access

Pyrogenic carbon emissions from the 2019 Yanchep wildfire

Patrick N Armstrong
Honours, Murdoch University
2022
pdf
Whole Thesis3.08 MBDownloadView
Open Access

Abstract

Fire is an important part of the carbon cycle and is widespread globally yet estimates of pyrogenic emissions are scarce for many vegetation types and often contain substantial uncertainty. Thus, there is a need to refine estimate methods and sample a broader array of ecosystems globally. This study uses a field-based approach to quantify carbon loss from a wildfire that burnt 12,300 ha of diverse heath and woodland vegetation in Yanchep, Western Australia, during Australia’s most severe fire year on record, known as ‘Black Summer’. Pre-existing field-based fuel load data were fused with 12 novel allometric equations, enabling empirical fuel estimates for 16 carbon pools across 5 vegetation types and 5 fire age classes. Burn-area was classified by vegetation type and fire history which permitted calculation of empirical fuels within the fire boundary. Combustion factors for each of the 16 carbon pools were adapted from published estimates and defined across 5 classes of fire severity. These data were combined with a pre-existing fire severity map generated from remotely sensed data and fed into a series of computations loosely based on the approach in Seiler and Crutzen (1980). Total pyrogenic emissions from the entire wildfire were estimated to be 42,587 tC, or 3.76 tC ha-1. Carbon loss corresponds to 33% of pre-fire fuel (11.32 t ha-1) with this consumption the result of high proportions of fine surface fuel and spatial coverage of severe fire conditions. Surface fuels were the major source of total fire emissions (44%) with shrub and canopy emissions increasing strongly with severity. This study presents an important step forward for the accurate quantification of fire-induced carbon loss in understudied Mediterranean-climate heath and woodlands.

Details

Metrics

110 File views/ downloads
109 Record Views
Logo image