Evidence of debris flow occurrence after wildfire in upland catchments of south-east Australia

Abstract

Numerous reports of “flash floods”, “mud torrents” and “landslides” in burnt landscapes of south-east Australia were only recently linked to debris flows and recognised as a significant process that warrant more detailed investigation. This paper provides a systematic documentation of high-magnitude erosion events after wildfire in south-east Australia, focusing on small (< 5 km 2), upland catchments in eastern Victoria that were burnt by wildfire between 2003 and 2009. The aims of the study were to i) collect and show evidence of debris flow occurrence after wildfire; ii) quantify erosion rates from debris flows and; iii) identify rainfall thresholds and key hydrological properties. The result showed that 13 out of the 16 recorded high-magnitude erosion events were runoff generated debris flows. These occurred in dry eucalypt forests burnt at high or very high severity in steep headwater catchments throughout the eastern uplands of Victoria. The debris flows were triggered by intense, short duration rainfall events ( I 30 35–59 mm h −1) with annual exceedance probability in the order of 20%. This is the first paper to document the occurrence of post-fire runoff generated debris flows in Australia, so the discussion draws on literature from the western USA, where a large body of research has been dedicated to evaluating the risk posed by post-fire debris flows and their role in landscape processes. Typical features common to both systems include low infiltration capacity of burnt catchments; widespread sheet erosion and levee lined rills on steep upper hillslopes; and severe channel erosion initiated in response to convergent flow in previously un-scoured drainage lines. The depth of sheet erosion on surveyed slopes in the upper catchments (4.6 ± 0.96 mm to 18.4 ± 2.7 mm) indicates that hillslope material provides an important source of sediment. The average channel entrainment rate of three debris flows ranged from 0.6 to 1.4 m 3 m −1. Runoff generated debris flows were not recorded in wet or damp forest types suggesting that this process is unlikely to operate in these forest environments. One isolated case of mass failure generated debris flow was recorded in wet forest. The outcome of the study indicates that runoff generated debris flows in dry eucalypt forest are an important process to be considered during post-fire risk assessment of hydrological hazards.