|Title||Testing woody fuel consumption models for application in Australian southern eucalypt forest fires |
|Publication Type||Journal Article |
|Year of Publication||2010 |
|Authors||Hollis, JJ, Matthews, S, Ottmar, RD, Prichard, SJ, Slijepcevic, A, Burrows, ND, Ward, B, Tolhurst, KG, Anderson, WR, Gould, JS |
|Journal||Forest Ecology and Management |
|Pagination||948 - 964 |
|Date Published||08/2010 |
|Abstract||Five models for the consumption of coarse woody debris or woody fuels with a diameter larger than 0.6 cm were assessed for application in Australian southern eucalypt forest fires including: CONSUME models for (1) activity fuels, (2) natural western woody and (3) natural southern woody fuels, (4) the BURNUP model and (5) the recommendation by the Australian National Carbon Accounting System which assumes 50% woody fuel consumption. These models were assessed using field data collected as part of the woody fuel consumption project (WFCP) in south-west Western Australia and northern-central Victoria. Three additional datasets were also sourced to increase variability in forest type, fuel complex and fire characteristics. These datasets comprised data from south-west Western Australia collected as part of Project Aquarius, the Warra Long Term Ecological Research site in Tasmania and Tumbarumba in south-eastern New South Wales. Combined the dataset represents a range of fire behaviour characteristic of prescribed burning conditions with a maximum fireline intensity of almost 4000 kW m−1.
Woody fuel consumption was found to be highly variable between sites ranging from 9.1% to 89.9%. Relationships between woody fuel consumption and the primary model drivers were weak (maximum R2 = 0.097). Model evaluation statistics were best for the National Carbon Accounting Systems assumption of 50% with a mean absolute error of 11.1% fuel consumption and minimal bias (0.12). Nonetheless, this assumption does not capture large deviations where woody fuel consumption has been particularly high or low. The BURNUP model yielded the largest level of error when used with natural fuels however its predictive capacity improved when used with large modified fuel loads resulting from clearcut operations.
|Short Title||Forest Ecology and Management |