Evapotranspiration, wildfire, and catchment water balance in south-east Australia

Download files
Access & Terms of Use
open access
Copyright: Webster, Evan
Altmetric
Abstract
Water accounting in remote heterogeneous landscapes is a largely uncertain process. Accurate estimates of actual evapotranspiration (ETa) are a vital component of the catchment water balance but elusive because of the confounding effects of wildfire, fluctuating precipitation patterns, and vegetation complexity. Spatial variability in data provision further exacerbates errors, as sparsely and irregularly distributed meteorological reference stations create uncertainty when interpolating meteorological data. To account for this, I tested the sensitivity of three surface energy balance (SEB) models to the preparation of meteorological data over a 35,167 km2 heterogeneous area in southeastern Australia. Two of the three SEB models were sensitive to uncertainty in meteorological inputs and the method of interpolation. The most effective SEB model for ETa estimation, the surface energy balance system, remained sensitive and vulnerable to error from input data. Therefore, I incorporated an iterative energy restraint to restrain its boundary conditions for SEB fluxes, improving accuracy and reducing sensitivity to bias and input error. I then used a spatially explicit temporal regression to implement a new post-wildfire impact and recovery vegetation response function, founded on the resistance and resilience to wildfire of forested ecosystems. This quantified the effect of wildfire and precipitation variability on vegetation in two contrasting areas in southeastern Australia with distinctly different vegetation, latitude, elevation, and wildfire histories. During post-wildfire vegetation recovery, the amount of precipitation immediately following a wildfire event was highly influential on the duration and intensity of vegetation regeneration in both areas. Establishing a balance between ETa, streamflow, and precipitation is challenging in heterogeneous catchments. By combining long-term catchment water balances with the Bowen Ratio, I customised the Priestley-Taylor parameter and used the Budyko framework to restrain soil evaporation in a process-based ETa model. This balanced ETa was then used to quantify the hydrological effect of a large catchment wildfire, indicating that increased fire severity drove substantial post-wildfire ETa reductions. This research quantified uncertainties and improved accuracy of SEB ETa estimates and developed a framework to assess the effect of wildfire on vegetation and ETa within heterogeneous catchments. This represents a major leap forward in mapping spatial and temporal changes in ETa, relative to wildfire and relevant to catchment water balances.
Persistent link to this record
Link to Publisher Version
Link to Open Access Version
Additional Link
Author(s)
Webster, Evan
Supervisor(s)
Kingsford, Richard
Ramp, Daniel
Keith, David
Sharma, Ashish
Creator(s)
Editor(s)
Translator(s)
Curator(s)
Designer(s)
Arranger(s)
Composer(s)
Recordist(s)
Conference Proceedings Editor(s)
Other Contributor(s)
Corporate/Industry Contributor(s)
Publication Year
2017
Resource Type
Thesis
Degree Type
PhD Doctorate
UNSW Faculty
Files
download public version.pdf 14.67 MB Adobe Portable Document Format
Related dataset(s)