Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Australie → Improved modelling of post fire hydro-geomorphic risks

University of Melbourne (2016)

Improved modelling of post fire hydro-geomorphic risks

Mason, Craig Ian

Titre : Improved modelling of post fire hydro-geomorphic risks

Auteur : Mason, Craig Ian

Université de soutenance : University of Melbourne

Grade : Master of Philosophy – Science 2016

Résumé
Hundreds of millions of people rely on forested water catchments for potable drinking water. However, rainfall subsequent to widespread wildfire can result in high concentrations of suspended sediment entering waterways, and eventually, water supply reservoirs. Calculation of the risk wildfire poses to water quality of supply reservoirs involves consideration of both probabilistic and deterministic processes which intersect at a wide range of scales and resolutions, in time and space. Two of the key requirements necessary to develop a suitable risk model are i) the generation of spatially distributed burn intensity data of future fires to inform subsequent erosion models, and ii) an improved understanding of the relationship between spatial scale and post fire hydro-geomorphic processes. This thesis addresses these two areas, to support the future development of a post fire hydro-geomorphic risk model. A probabilistic wildfire burn area simulation model is developed and presented which delivers the required spatially distributed burn intensity data, while also replicating the fire frequency and magnitude of the local fire regime. The model relies on historical fire records and the adaption of an existing fire behaviour simulator to produce the desired output data. These data can then parameterise deterministic hydrological and erosion process sub-models to determine local post fire water quality risk. Further development of the wildfire model will allow more accurate simulation of the local fire regime.
The relationship between spatial scale and post fire hydro-geomorphic processes is poorly understood, which prevents further development of a post fire hydro-geomorphic risk model. This thesis includes an analysis conducted to improve understanding of wildfire effects on peak discharge at the large catchment scale. Pre and post wildfire rainfall and runoff data from six large catchments in Australia’s southeast upland forests were analysed to determine whether wildfire causes significant increases in peak discharge. No significant change in peak discharge following wildfire at the large catchment scale was detected, however, substantial data quality issues clouded clear conclusions. This analysis concurs with the few local studies available, implying that at large catchment scales, post fire suspended sediment loads measured in stream networks after large volume rainfall events may come from near-channel, channel bank, or in-channel sources. This may also suggest that large volume rainfall events may not need to be considered when modelling hillslope erosion processes, possibly reducing the necessary complexity of an overall risk model.
This thesis provides additional tools and knowledge to support future modelling of post fire hydro-geomorphic risks. A variety of other processes still require investigation and sub-model representation before the range of post fire suspended sediment loads delivered to water supply reservoirs by the hydro-geomorphic system can be reliably calculated.

Présentation

Version intégrale (3,07 Mb)

Page publiée le 8 novembre 2019