Bushfire, Storms and Soil Erosion
Dr. Andrew Hammond and Dr. Michael C. Clarke
Share this page:
Fire and storm events have had a significant influence on the type, structure and conservation of soils in eastern Australia. Fire of anthropogenic and natural origins has been the cause of great changes in the flora and soil fauna of the region. Storm events are related to seasons and vary in terms of intensity on a quasi-seasonal basis in some sub-regions. They are also influenced by the La Nina/El Nino cycle.
The interaction of fire and storm events is examined in this paper. The intensity of storm events of high or low precipitation intensity will have markedly different effects on the soils, where the storm is immediately preceded by fire. Where fire has occurred in close time sequence with storm (precipitation) events erosion can be enhanced and hence soil profiles permanently altered.
Fire management includes control burning as a tool. Fire management that uses control burning should also take into account the soil condition of the areas designated to be burnt, the immediate storm and precipitation history, and the likely (expected) meteorological conditions for the season. Control burning should be minimised where severe storm events are likely and where they occur without low-level precipitation being a likely precursor. The condition of the underlying soil in an area to be burnt should be taken into account before burning is undertaken.
The effect of having to consider soil condition and the probability of extreme storm events, subsequent to controlled burning, will reduce the window of opportunity for controlled burning. This will have the effect of requiring additional and more flexible resources for controlled burning and other fuel reduction procedures, being made available.
The interaction of fire and storm events has had a significant influence upon the type, structure and conservation of eastern Australian soils. Bushfire of anthropogenic and natural origin has seriously changed the flora and fauna of Australia (Flannery 1995). Storm events, and rainfall in general, are related to seasons and vary in terms of intensity on a quasi-seasonal basis in some sub-regions. They are influenced by La Nina/El Nino cycle, which in turn has a gross influence on the timing, frequency and intensity of storms.
Storm events of either high or low precipitation intensity, occurring in specific precipitation sequences following fire. have markedly different effects upon soils. Where fire has occurred in close time sequence with storm (precipitation) events erosion can be enhanced and soil profiles permanently altered (Greene et al., 1990 & Kinnell et al., 1990)
Severe fire events, followed immediately by heavy rain, are costly for soil in terms of erosion. The loss of ground cover, plant foliage and the reduction in soil cryptogams (the micro-life forms that help maintain soil cohesion) make soil liable for degradation. Where soil cover is lost the impact of raindrops will be more damaging, since there will be less opportunity for the impact to be mitigated by droplet dispersal on foliage and impact dampening by ground cover. Soil will be left exposed and vulnerable to soil erosion.
Where a fire event is coupled with drought root and soil flora, and fauna activity, will be reduced or forced into inactivity. If heavy (and persistent) rain occurs without shower precursors, then raindrops will penetrate the exposed soil, dislodge soil particles, allowing the aggregated precipitation to carry the soil, and the fire damaged cover, away. If shower precursors occur before heavy rain, there will be a possibility for the generation of new foliage, and the laying down and consolidation of ground cover. The sequence of showers and rainfall events is hard to predict, but prescribed fire events (control burns) are planned.
Fire management strategies need to consider: the soil condition of the areas designated to be burnt, the storm and precipitation history, and the likely (expected) meteorological conditions for the season. Control burning should be minimised where severe storm events are likely and where they occur without low-level precipitation being a likely precursor.
The effect of considering soil condition, together with the probability of extreme storm events subsequent to controlled burning, will reduce the 'window of opportunity' for burning. Consequently, additional and more flexible resources for controlled burning and other fuel reduction procedures, need to be made available. This paper examines some consequences of and strategies for managing fire using controlled burning, with these added restrictions.
2. RAINDROP SIZE AND IMPACT DAMAGE POTENTIAL
Raindrop size plays a pivotal role in detaching individual soil particles making them liable for subsequent movement by water. Thus erosion begins.
Anthoni (2002) states that the damage done by raindrops is related to raindrop size and 'the energy of a moving object is equal to its mass multiplied by its speed squared: e = mv2. As water droplets grow in size, both their speed and mass increase. The mass of a 5 mm raindrop is 5x5x5 = 125 times that of a 1 mm drop and its 'terminal' speed doubles, resulting in a destructive energy 500 times larger!' Thus if an area has a minimum of 95% foliage and ground cover (e.g. forests and grasslands) then raindrop impact damage would be light. In heavily wooded areas the average terminal velocity of the droplets would be low, since some raindrops would have undergone fractionation upon collision with foliage. For grasslands, the vertical velocity of remnant drops would be close to zero if there is understory foliage that intercepts the raindrops and is close to ground level. The flow of water to the soil in this case would be largely stem-flow.
Furthermore, Anthoni states 'the destructive power of rain increases dramatically as the rainstorm produces larger drops, which is relatively rare. But when it occurs, its effect is profoundly destructive. In the past ten years, starting around 1987, rains have become heavier everywhere in the world, and with it, erosion from raindrop impact.' If this proposition of Anthoni on climate change is correct, then our exposed soils are in for a punishing time.
Outside the realm of recent climate change conjecture (being naturally or anthropogenically induced) and the alarmist conclusions of those who support the hypothesis, raindrop size is an important factor in soil erosion (Rose 2003). Large raindrops, of say 5mm diameter, will cause significantly more soil particle movement than small drops of under a millimetre.
Raindrop sizes under natural conditions vary between a fraction of a millimetre to around 5mm (Overton, 1996). If the average raindrop size is around 5mm (termed gravity drops), then enhanced erosion will result (Moss & Green, 1987). Since gravity drops form on leaf surfaces and require around 11.2 metres of fall to reach terminal velocity, drops from a forest canopy falling onto a fire denuded soil surface will be significant enhancers of erosion.
3. SOIL NATURE AND FIRE EVENTS
Where all foliage, leaf litter and organic material have been removed by fire, soil is liable to be eroded by the raindrop impact breaking up and dislodging the surface material. In a situation that has a severely burnt understory, but relatively untouched canopy, gravity drops may be a significant cause of soil particle displacement. In either case the protection of soil against erosion, as provided by foliage and its remnants, will be missing.
In severe fire events an increase in soil hydrophobicity may occur resulting in hydraulic conductivity being lowered and runoff from rainfall being doubled (Moss & Green, 1987, and Elliot et al., 1991). The loss of soil from severe fire affected areas approaches the loss that could be expected through clearing by forestry activities.
DeBano (2000) states, 'the combination of fuel combustion and heat transfer during wildfires produces steep temperature gradients in the surface layers of the mineral soil. Heat produced during combustion of litter and above-ground fuels vaporizes organic substances which are moved downward into the underlying mineral soil where they condense in the cooler underlying soil layers, forming a distinct water-repellent (hydrophobic) layer below and parallel to the soil surface'.
Kinnell et al. (1990) state that this change in hydrophobicity may result in the destruction of the cryptogam cover down to 10 mm. Changes in the chemical sub-layer may thus reflect damage done to the cryptogam cover.
Soil condition should be monitored both prior to and post burning as it is a significant factor in erosion control. The moisture content of soil, as indicated by the Byram and Keetch Drought Index (BKDI) (Keetch and Byram, 1968), affects the moisture content of the overlying fuel.
The use of the BKDI however is not currently intended for assessing the likelihood of erosion following fire events.
4. FIRE MANAGEMENT AND SOIL CONDITION
Fire management is the protection of life, property and the environment and has many contributing/controlling factors. The primary management consideration being the minimisation of fuel that can be a hazard. Fuel reduction includes the removal of all significant flora from an area (it is hard to set fire to a mown football field), the thinning of fuel and/or the removal or reduction of plant species with high volatile contents. The practice of controlled burning is the land/fire management practice that this paper concentrates on.
The use of controlled burning such that soils are affected as little as possible and erosion is minimised is considered to be essential. Some considerations for good fire management with respect to the preservation/ conservation of soils are:
· The ability to target specific fuels for removal,
· The ability to avoid the removal of organic material required for soil protection and plant growth,
· The condition (e.g. hydrophobicity) of the soil prior to burning, and
· The expected climatic conditions during and immediately after the fire event.
Conducting burn-offs such that understory fuel is effectively removed while maintaining soil cover is the first objective. Overton (1996) of the NSW Bushfire Service recommends that burn-offs in open eucalypt forests be limited to a maximum flame height of 1.5m and a fire intensity of <500kW/m, to prevent excessive damage to the canopy and the groundcover.
A secondary objective of good fire management is the avoidance of excessive understory scorching that exposes soil but leaves significant canopy cover should also be avoided. The setting of fire that leaves neither soil cover, viable trees nor cryptogam activity is the worst scenario, and is the failure of the management programme. The next question posed is when to burn?
4.1 The Use of Erosivity and Historic Fire Season Rainfall Data in Control Burn Planning
Lu and Yu (2002) have used the Revised Universal Soil Loss Equation (RUSLE) to examine soil loss through intense rainfall events. They have used the rainfall erosivity (R) factor for indicating the potential for water erosion for numerous Australian sites. The R factor has the units of MJ.mm/(ha.h.year) and for Brisbane is recorded as being of the order of 4901 using a zero threshold.
The R factor is a combination of kinetic energy (E) and rainfall intensity I30 (as recorded by Pluviographs and rain gauges). The data, Figure 1, has further been broken down into monthly figures as percentages.
Figure 1. Monthly R-factor distributions for Brisbane - 20 year data (Lu & Yu, 2002)
High erosivity events can be seen to occur from late spring to late autumn. The months April to July have successively decreasing erosivity readings. June and July have low total precipitation whilst April and May have good overall precipitation. The desired situation of distributed rainfall without a high propensity for intense rainfall events is the month of May.
By combining data from Figure 1, with data on bushfire occurrence as supplied by Walker et al. (1986) an annual pattern of fire occurrence and periods of good rainfall with low probability of intense precipitation events can be developed, Figure 2. In S. E. Queensland the period of high bushfire occurrence is between weeks 30 and 45, late winter to late spring, of the year. During this period control burning should not be considered. Likewise, during the summer months the probability of intense rainfall events is high, whilst total precipitation is also high.
Figure 2. The Brisbane Fire Season superimposed with precipitation and Pluviograph data
From Figure 2, May is the ideal month for controlled burning in the Brisbane region with the months of April and June being 'shoulder' months to help with resource planning.
If controlled burning was to be extended into the winter, then the problem of reduced plant re-growth will be encountered. During June and July (winter) Brisbane night temperatures are around 10°C. As this is the annual 'drought' period, burnt foliage will not regenerate quickly.
5. RISK MANAGEMENT FOR EROSION CONTROL AND CONTROL BURNING
Burn Plans (BPs) and Operational Fuel Management Plans (OFMPs), as produced by those involved in controlled burning, are in essence Risk Management Plans. They primarily look at the likelihood of fire that is set for controlled burning being containable to those areas chosen for burning. The second consideration in BPs and OFMPs is the degree to which the burn-off will reduce the fuel load and hence the fire hazard in the control burning area and its surrounds.
Within the State Forests of New South Wales Model Burning Plan (Berry, 2002), there is consideration of the likelihood for soil erosion presented under the heading 'Environmental Prescriptions', and sub-heading, 'Soil, Water & Aquatic Habitat, with 'rainfall erosivity -preferred months of burn .?'. In the OFMP for the Hastings area of northern New South Wales, no specific mention of fire and rainfall erosivity is mentioned in the State Forest's operational fuel management plan - environmental impact assessment provisions. Thus in one government instrumentality there are differing levels of importance given to fire induced soil erosion. This lack of conformity in Risk Management practice needs be examined and rectified.
The development of regional and sub-regional burning plans that incorporate erosion considerations that are based on data from rainfall, rainfall intensity as well as the historical occurrence of fire, and the immediate physical circumstances of a location (temperature, wind speed, BKDI index and fuel load) would be a useful incorporation into burn plans and operational fuel management plans.
5.1 Resources, Risk Management and Control Burning
The more factors that are brought into consideration in a BPs or OFMPs, the narrower the window for controlled burning will be. This will have significant impacts upon the use of resources and costs involved in controlled burning. It will further have the effect of limiting the use of volunteers for controlled burning, since personnel will be needed 'on demand' and not on a convenience basis.
In the Brisbane example, the limiting of controlled burning to late autumn, will necessitate the availability of fire personnel on a whole week basis, since for some days in any period it will not be possible to burn-off because of the prevailing weather conditions.
If a broader regional view is taken, where a number of regions that have different optimal controlled burning periods are included in a 'super fire control region' then resources (personnel and appliances) may be able to be better utilised, that is to be shared.
Volunteerism has been a core of the rural fire and bushfire services. To be able to burn on a demand basis will mean that the involvement of bushfire brigades will be lessened, and that employed personnel will need to be made available. This will have two primary effects: the cost of control burning will increase, and that the role of volunteers in the rural fire services will need to be redefined. This cost is however small compared to the financial losses and social problems experienced by a community devastated by bushfire.
6. FUTURE RESEARCH DIRECTIONS
One question that may be useful to answer is can the drought index, as developed by Byram and Keetch (1968), be utilised in providing information of the erosivity potential of soil following fire. This should be undertaken prior to a controlled burn and possible scenarios formulated.
Some factors to be examined in the investigation would be:
· does the BKDI give a good indication of the health of soil cryptogams before burning,
· what is the effect of various control burning strategies on soil cryptogams' viability,
· can the BKDI give an indication of the propensity for erosion, and
· is the alternative k and R factors in the revised Universal Soil Loss Equation more relevant indicators of potential soil loss.
US research (originating in South Carolina) (Keetch-Byram) has indicated that a BKDI of 37.5 to 75 (using the metric scale of 0 - 200) 'will allow scattered patches of surface litter to remain in damp areas following a fire, and the organic layer (including soil cryptogams?) remains basically undisturbed'. In the Hastings area the maximum BKDI for a control burn was given as 70 in their OFMP. It should also be noted that the US research found that for a BKDI of between 75 to 125, 'fire consumes most surface litter along with a significant loss in organic soil material. Site preparation burns expose mineral soil, producing areas causing erosion problems.'
The combination of using the BKDI (or similar index) plus meteorological data on the likelihood of specific rainfall events may well hold the answers to safe control burning-off in terms of erosion limitation.
Another area of research interest is the matching of erosivity data with historical seasonal fire pattern data. Where data exists down to a micro-climate scale, sub-regional controlled burn 'safe periods' could be defined in terms of potential erosivity.
By taking into account the climatic conditions that can be reasonably forecasted during and following a period of controlled burning the window of opportunity for controlled burning will be reduced. The optimal situation is to schedule controlled burning during periods of light rain, with minimal disturbance of decomposing leaf-litter and cryptogams.
In terms of the use of controlled burning in fire management, the added restriction of being more aware of the seasonal climatic conditions as well as soil condition will reduce the window of opportunity for controlled burns. This will have the effect of increasing the cost of fire management, since volunteerism is now such a big factor in current fire management plans. Volunteers will need to be replaced with 'on-tap' crews who will respond primarily to control burning opportunities and not haphazard availability.
A reduction in the window for controlled burning will unfortunately occur once soil and extended climate considerations are included in fire management plans. To make control burning feasible as a forestry management tool under the extended control criteria, additional resources will be required to take advantage of those windows of opportunity.
The authors wish to acknowledge the contribution of: Professor Calvin Rose, Griffith University for his assistance on raindrop size, Mr. Michael Berry, Cumberland State Forest, NSW, for the Burning Plans and Operational Fuel Management Plans, and Dr. Bofu Yu, Griffith University for information on rainfall erosivity.
Anthoni J. F. (2000) - Seafriends. Soil, Erosion and Conservation. www.seafriends.org.nz/enviro/soil/erosion
Berry M. (2000) Personal Communication from, Manager, Cumberland State Forest, NSW. Burning Plan Templates.
De.Bano L. F. (2000) US Department of Agriculture, Forest Service Proceedings pp 307 - 310,
Elliot W J, Foster G R, Elliot A V: (1991) Soil erosion: processes, impacts and prediction.
Ed. Lal & Pierce Soil management for sustainability, CRC Lewis Publishers, Boca Raton.
Flannery, Tim F. (1995) The future eaters: an ecological history of the Australasian lands and people. Sydney : Reed Books.
Greene, R. S.B., Chartres, C.J., and Hodgkinson, K.C., (1990) The effects of fire on the soil in a degraded semi-arid woodland. I. Cryptogam cover and physical and micromorphological properties. Australian Journal of Soil Research 28: 755-777.
Keetch, J.J. and Byram, (1968.) G.M.. A drought index for forest fire control. US Department of Agriculture.Forest Service. Paper Se-38,
Keetch-Byram Drought Index Description, Dept. Of Natural Resources South Carolina. www.dnr.state.sc.us/climate/sco/drought/keetch
Kinnell, P.I.A.., Chartres, C.J., and Watson, C. L. (1990) The effects of fire on the soil in a degraded semi-arid woodland. II. Susceptability of the Soil to erosion by Shallow Rain-impacted Flow. Australian Journal of Soil Research 28: 779-794.
Lu H. and Yu B. (2002) Spatial and seasonal distribution of rainfall erosivity in Australia. Aust. J. Soil Res., 40, pp 887 - 901.
Moss A. J. & Green T. W. (1987) Erosive effects of large water drops (gravity drops) that fall from plants. Aust. J. Soil Resources. Vol. 25, pp 9 - 20.
Overton, F. (1996) Fire Fighting Management and Techniques. Inkata Press, Melbourne.
Rose C. Personal Communication, 2003.
Walker J., Rainson R. J. and Khanna P.K. (1986) Fire, Chapter 8, Australian Soils, The Human Impact. University of Queensland Press.
You are welcome to quote up to a maximum of three paragraphs from the above white paper, on condition that you include attribution to this website, as follows:
SOURCE: M.E.T.T.S. Pty. Ltd. Website http://www.metts.com.au
Share this page: