The vegetation of the Sydney Basin, Australia, is highly flammable and subject to a wide range of fire regimes. Sclerophyllous shrubs and sedges are common and in some vegetation types up to 70 % of fuel consumed during a fire can be live. Research into fire behaviour and fuel dynamics has been minimal. To address this issue this thesis investigated the principal factor affecting the ease of ignition and rate of combustion of individual fuel particles and fuel beds in bushfires: dead fine fuel moisture (FFM). Two common Sydney Basin vegetation types, eucalypt woodland and heathland, each with a history of problematic fire management, were measured in the field for diurnal fluctuations in FFM following rain, under conditions similar to when prescribed burns are conducted. The FFM components of current operational fire behaviour models were found to be inadequate for predictions of FFM and fire behaviour under these conditions. The equilibrium moisture content (EMC) of five fuel types from the field site was investigated in a laboratory study. An existing function describing EMC as a function of temperature and relative humidity was evaluated and found to be very accurate for these fuels. Two FFM predictive models incorporating this function were evaluated on the field data and the laboratory results were shown to be applicable to the estimation of FFM in the field. One model gave very accurate predictions of FFM below fibre saturation point, but its accuracy was reduced when screen level conditions were used instead of those measured at fuel level. A recent process-based model that accounts for rainfall showed promise for predicting when fuel is < 25 % FFM. Systematic problems with the radiation budget of this model reduced the accuracy of predictions and further refinement is required. Live fine fuel moisture content (LFMC) of common heathland shrubs and sedge was investigated over two years and found to be both seasonal and influenced by phenology. LFMC minima occurred in late winter and spring (August to October), and maxima were in summer (December to February) when new growth was recorded. The dominant near-surface fuel in mature heath was sedge. It was found to have little seasonal variation in its??? percentage dead but the percentage dead maxima occured at the same time as the LFMC minima of shrubs and sedge in both years. Simple instantaneous models for duff moisture content in woodland and heathland and LFMC and the percentage dead sedge in heathland were developed. The information gained by this study will form the basis for future development of fuel moisture models for prescribed burning guidelines and fire spread models specific to the vegetation communities of the Sydney Basin.

The vegetation of the Sydney Basin, Australia, is highly flammable and subject to a wide range of fire regimes. Sclerophyllous shrubs and sedges are common and in some vegetation types up to 70 % of fuel consumed during a fire can be live. Research into fire behaviour and fuel dynamics has been minimal. To address this issue this thesis investigated the principal factor affecting the ease of ignition and rate of combustion of individual fuel particles and fuel beds in bushfires: dead fine fuel moisture (FFM). Two common Sydney Basin vegetation types, eucalypt woodland and heathland, each with a history of problematic fire management, were measured in the field for diurnal fluctuations in FFM following rain, under conditions similar to when prescribed burns are conducted. The FFM components of current operational fire behaviour models were found to be inadequate for predictions of FFM and fire behaviour under these conditions. The equilibrium moisture content (EMC) of five fuel types from the field site was investigated in a laboratory study. An existing function describing EMC as a function of temperature and relative humidity was evaluated and found to be very accurate for these fuels. Two FFM predictive models incorporating this function were evaluated on the field data and the laboratory results were shown to be applicable to the estimation of FFM in the field. One model gave very accurate predictions of FFM below fibre saturation point, but its accuracy was reduced when screen level conditions were used instead of those measured at fuel level. A recent process-based model that accounts for rainfall showed promise for predicting when fuel is < 25 % FFM. Systematic problems with the radiation budget of this model reduced the accuracy of predictions and further refinement is required. Live fine fuel moisture content (LFMC) of common heathland shrubs and sedge was investigated over two years and found to be both seasonal and influenced by phenology. LFMC minima occurred in late winter and spring (August to October), and maxima were in summer (December to February) when new growth was recorded. The dominant near-surface fuel in mature heath was sedge. It was found to have little seasonal variation in its??? percentage dead but the percentage dead maxima occured at the same time as the LFMC minima of shrubs and sedge in both years. Simple instantaneous models for duff moisture content in woodland and heathland and LFMC and the percentage dead sedge in heathland were developed. The information gained by this study will form the basis for future development of fuel moisture models for prescribed burning guidelines and fire spread models specific to the vegetation communities of the Sydney Basin.