Australian wildfires and risks of increased erosion rates

(Updated 10th Feb to include latest casualty numbers)

The extraordinary wildfires in Australia are dominating the headlines in the UK, half a world away. Wildfires are quite common events, but the number of fatalities that this particular episode has caused is really quite unusual. Below in Figure 1 I have plotted the recorded worldwide recorded number of deaths from wildfires for the period since 1980, using data from the CRED EM-DAT database . I have added the (updated to) 173 reported deaths from this event so far as an extra column, although note that reports suggest that this total may rise substantially:

Figure 1: Global numbers of reported fatalities from widlfires, based upon the CRED EM-DAT database. The 2009 value is the reported number of deaths from the Austrlian wildfires.

The average annual global total number of deaths is 59.5 fatalities per annum. Care is needed in the interpretation of the above as CRED only record events that kill ten or more people, thus these values consistently underestimate the true toll, but nonetheless the unusual impact of these events is clear.

In the context of this blog it is also interesting to think through the likely long term impact of these fires in terms of erosion and landslides. A recent paper by Smith and Dragovitch (2008) looked at the long term consequences of wildfires in SE. Australia. These two researchers have published extensively on sediment production and erosion in Australia, so have a strong pedigree.

The study focused on a fire that occurred in January 2003 during a drought in the Snowy Mountains near to Thredbo (Fig. 2), a sub-alpine environment. The study used erosion pins to monitor surface level change on both burnt and unburnt hillslopes over a period of 795 days after the fire.

Figure 2: Google Earth perspective image of the area around Thredbo, the location of the study reported by Smith and Dragovich (2008)

The study showed that after the fire the areas that had been burnt suffered a net loss of soil representing an average of 3.8 mm of material, with the most intense erosion occurring on the lower slopes (Fig. 3). On the other hand, the unburnt areas saw a net accumulation of soil of an average of 2.6 mm, again with the greatest accumulation at the lower slopes.

Figure 3: Mean net soil loss and gain for burnt and unburnt areas as reported by Smith and Dragovich (2008).

Thus, the burnt areas clearly suffered a net loss of material in the aftermath of the fires. The study showed that this loss of soil declined with time after the fire, with a slight increase again during snow melt, presumably as vegetation re-established. However, these values are perhaps surprisingly low compared with those recorded in other environments, especially in N. America, given the steep slope angles seen in Figure 2. Interestingly, Shakesby et al. (2007), who studied post-fire erosion in Eucalyptus forests in SE. Australia, came to similar conclusions, stating that "except under extreme post-fire rainfall conditions, present-day wildfires affecting south-east Australia seem to be less potent in geomorphological terms than might be expected given the severity and frequency of the wildfires". They attribute this to the rapid rate of plant growth in the aftermath of fires plus the resistance of the soil to erosion.

The conclusion is therefore that although the fires have devastated vast areas, and made thousands homeless, there should not be a serious increase in erosion in the burnt areas. This will help greatly in the post-fire recovery of the burnt areas.

References:
H SMITH, D DRAGOVICH (2008). Post-fire hillslope erosion response in a sub-alpine environment, south-eastern Australia CATENA, 73 (3), 274-285 DOI: 10.1016/j.catena.2007.11.003

R SHAKESBY, P WALLBRINK, S DOERR, P ENGLISH, C CHAFER, G HUMPHREYS, W BLAKE, K TOMKINS (2007). Distinctiveness of wildfire effects on soil erosion in south-east Australian eucalypt forests assessed in a global context Forest Ecology and Management, 238 (1-3), 347-364 DOI: 10.1016/j.foreco.2006.10.029