Christopher Lucas

Prediction of the probability of large fires in the Sydney region of south-eastern Australia using fire weather

The probability of large-fire (≥1000 ha) ignition days, in the Sydney region, was examined using historical records. Relative influences of the ambient and drought components of the Forest Fire Danger Index (FFDI) on large fire ignition probability were explored using Bayesian logistic regression. The preferred models for two areas (Blue Mountains and Central Coast) were composed of the sum of FFDI (Drought Factor, DF = 1) (ambient component) and DF as predictors. Both drought and ambient weather positively affected the chance of large fire ignitions, with large fires more probable on the Central Coast than in the Blue Mountains. The preferred, additive combination of drought and ambient weather had a marked threshold effect on large-fire ignition and total area burned in both localities. This may be due to a landscape-scale increase in the connectivity of available fuel at high values of the index. Higher probability of large fires on the Central Coast may be due to more subdued terrain or higher population density and ignitions. Climate scenarios for 2050 yielded predictions of a 20–84% increase in potential large-fire ignitions days, using the preferred model.

Meteorological conditions and wildfire-related houseloss in Australia

Wildland fires or bushfires occurring under very severe weather conditions are likely to be destructive to infrastructure. This paper reports an analysis of the statistical relationship between house loss and the fire weather under which it occurred. A dataset was derived from 54 bushfires that occurred in Australia between 1957 and 2009, which resulted in the destruction of 8256 houses. The dataset was statistically compared with relevant local meteorological conditions, and a standardised calculation of the McArthur Forest Fire Danger Index (FFDI) applied. The analysis highlights how house loss statistics in Australia are dominated by a few iconic events that have occurred during very intense fire weather with the majority of losses occurring on days when the FFDI exceeds 100. Virtually all of the house loss has occurred above the 99.5th percentile level in the distribution of daily FFDI for each of the regions considered. Regulatory tools will need to focus on the most appropriate fire weather potential of a local area in order to ensure that infrastructure is adequately designed. In Australia, little house loss has occurred on days where the FFDI did not exceed 50, suggesting that historic building practices may be maintained in regions where this level is not likely to be exceeded.

Interannual variations of area burnt in Tasmanian bushfires: relationships with climate and predictability

The area burnt each summer in Tasmania is related to coincident (summer) climate variables, especially the total summer rainfall. The relationship with temperature is weaker and largely reflects the relationship between rainfall and temperature. As the El Nino–Southern Oscillation is known to be related to Australian rainfall and simple indices of this phenomenon form the basis of the operational seasonal climate forecast scheme used in Australia, it is not surprising that indices of the El Nino–Southern Oscillation can also, it appears, provide a potentially useful forecast system for Tasmanian bushfire extent. In particular, sea surface temperatures in the Coral Sea during winter are correlated with the area burnt in the following summer. The effect of summer rainfall on the area burnt each year suggests that global warming may not simply lead to increased burning, contrasting with the situation in other parts of the globe. A weak, long-term decline in area burnt appears to be due to a weak increase in summer rainfall.

Regional characteristics of tropical expansion and the role of climate variability

Radiosonde-based tropical expansion rate estimates for six continental-centered regions of the globe are discussed. New results from the Northern Hemisphere are presented, complementing previous Southern Hemisphere (SH) results using an identical methodology. Expansion rates are largest over Asia and Australia-New Zealand (ANZ). Other regions show more modest rates of expansion, and there is no statistically significant expansion over North America. In the hemispheric average, expansion rates are slightly larger in the SH. This asymmetry, although not statistically significant, is consistent with ozone depletion acting as a codriver of tropical expansion since 1979. The picture of regional expansion here is different from that in other studies or that derived solely from reanalyses. The relationships between tropical expansion and modes of climate variability are explored using partial regression techniques. Interannually, a relationship is seen with the El Nino–Southern Oscillation (ENSO) across most of the globe. Regionally, the Pacific Decadal Oscillation (PDO) has a significant impact over Asia, while the Southern Annular Mode (SAM) affects expansion over ANZ. Considering the longer-term changes in the climate variability indices since 1979, it suggests that the PDO may account for up to 50% of the observed tropical expansion over Asia, while SAM may account for 20–30% of the SH expansion, particularly over ANZ. Decadal changes in ENSO may account for a further 20–30% of the global and regional trends. Accounting for the effect of climate variability helps to reconcile the differences in observations and model-based simulations of tropical expansion.

Estimation of optimal dispersion model source parameters using satellite detections of volcanic ash

In this paper we demonstrate how parameters describing the geometry of the volcanic ash source for a particular volcanic ash dispersion model (HYSPLIT) may be inferred by the use of satellite data and multiple trial simulations. The areas of space likely to be contaminated by ash are identified with the aid of various remote sensing techniques and polygons are drawn around these areas as they would be in an operational setting. Dispersion model simulations are initialized either by a cylindrical source or a specified ash distribution depending on the context. Parameters of interest such as the base and top height, diameter, and optimal release time of the cylindrical source or the height of the specified ash distribution are inferred by forming a parameter grid and running multiple simulations for each parameter grid-point value. Optimal values of the parameter values are identified by calculating spatial correlations between the model simulations and observations. We demonstrate that the methodology can be used to correctly infer various model parameters and improves volcanic ash forecasts in various eruption case studies.

Toward quantitative forecasts of volcanic ash dispersal: Using satellite retrievals for optimal estimation of source terms

The provision of reliable quantitative forecasts for volcanic ash, such as ash mass load fields, is challenging because ash emission characteristics at the volcanic source are poorly understood. In this paper we show how satellite retrievals of volcanic ash mass load may be used to estimate source terms for dispersion models. The source terms comprise the spatial, temporal, and particle size distribution of mass flux at the ash source. We approach the problem by specifying general functional forms for these quantities that are dependent on a limited number of parameters. Numerous trial dispersion model simulations are then run, each corresponding to a particular configuration of possible source term parameters, with the resulting simulated mass load matched against the satellite retrieved mass load. The parameter values leading to best matches between simulations and satellite retrievals are then used to provide optimal forecasts of volcanic ash mass load distribution. We use several case studies to demonstrate the efficacy of this approach in improving forecasts of ash mass load with the HYSPLIT dispersion model.