Fire in the North End: Some Results
Significant firescars were mapped each year in the Kimberley region in north western Australia between 1993 and 1997.
The worst year was 1997, when 31% was burnt. The lowest annual burning occurred in 1993, when less than 4% was burnt. The figures are similar in the Northern Territory. Between March and November 1997, the total number of hot spots reported for the Kimberley, were 5759 and in the Northern Territory 6337.
The breakdown into monthly statistics (Figure 5) indicates little burning before April, when prescribed burning usually begins. Towards the end of the dry season – August through to December – the number of fires and area burnt increases significantly. The fires later in the year also tend to cover a larger area.
The high number of hot spots does not necessarily indicate the number of individual fires. During the drier months the fires are large and burn continuously for weeks and even months. They burn on many fronts and each fire front can have numerous hot spots associated with it
A GIS is used to produce maps of the frequency and extent of biomass. The fire history vectors can be combined with a vegetation map to determine the dominant vegetation type in each area of interest. The areas that were burnt once in the four years were likely to be shrub steppe, or sparse low tree steppe. The areas that burnt two years out of the four were mainly the shrub steppe, scattered shrubs or high grassland savanna. Areas that burnt three out of the four years were the high grass savanna, tall bunch grass, and the fringes of the tree steppe. Areas that burn every year consist of high grass savanna.
The frequency and extent of fire closely follow the seasonal monsoonal patterns and the vegetation type. In 1991 and 1992 there was a severe drought in the Kimberley limiting fuel load build up. In 1993 the monsoon began two months late with the NDVI indicating that the build up of green vegetation cover did not begin until March. It remained unusually green into September and October.
It therefore appears that far less of the Kimberley burnt in 1993 due to the previous two years of drought. This limited the build-up of fuel. The late monsoon caused the vegetation to be insufficiently cured for significant burning to occur before the onset of the next wet season.
Further monitoring over a number of years will help to further understand the patterns of burning and provide more information for these interpretations, so that a prediction of fire based on the previous monsoon, vegetation type and fire history can be developed. We have an archive of NOAA data from 1982, so it is feasible to map fires for the last 15 years.
(This page last modified on 1 December 1998)