Impacts of Wildfire on Water Flows from Forested Catchments
Water Used by Vegetation
All plants evaporate water through their leaves. This water is extracted from the soil root zone, and the rate of evaporation depends on the weather, the available soil moisture, and the total area of leaves in the vegetation (trees and understorey). There are differences between various forest types, but basically different forests have evolved to make optimum use of the available rainfall to ensure their survival. Streamflow in drier periods is the "left-over rainfall" that passes beyond the root zone and exudes into the stream from boggy areas and the water table next to the stream. In storms, water runoff also occurs where the rainfall is intense enough to exceed the capacity of the soil to absorb it, or where the soil is already saturated. This runoff results in rapid increases in streamflow, or floods during major storms.
For example, during an average year at a south eastern Australian catchment where the annual rainfall is 1000 mm, the forest canopy may intercept and evaporate 150 mm of the rainfall before it reaches the ground. The forest may consume a further 750 mm by plant transpiration, leaving only 100 mm to appear as streamflow (this is equivalent to a water yield of 1 megalitre per hectare). Of this 100 mm, 80 mm may occur as short-term runoff during storms, while the remaining 20 mm occurs as sustained dry-weather flow or "baseflow".
These proportions depend on forest type (e.g. pines versus eucalypts), soil type and the local climate. Pine forests have denser canopies than most native forests, so they can intercept more water (about 250 mm in the example above), especially if the dominant rainfall is low intensity and uniformly distributed throughout the year. These factors alone cause large differences in streamflow yields from different forest types. On the other hand, soil types affect the fraction of total flow that appears as baseflow (described later). Baseflow can be as high as 50% of total flow in catchments with deep, well-structured soils.
In all forests, the vegetation transpires water at a rate that depends on solar radiation and atmospheric conditions. But when the moisture available in the root zone becomes seriously depleted, plants respond by limiting water use through their leaves. Depending on the severity of the drought, leaf fall and tree death can occur. The processes differ between tree species, and are not well understood even for the major forest types. As a result, our understanding of water yield from forested catchments and the impacts of changing the vegetation cover from grassland to forest or changing the forest type are currently areas of active research.
|evapotranspiration||View Frequently Asked Questions||View Bibliography|
|eucalypts||View Frequently Asked Questions||View Bibliography|
|pines||View Frequently Asked Questions||View Bibliography|
|water yield||View Frequently Asked Questions||View Bibliography|
Short-term Changes in Streamflow
Fire immediately destroys the ability of vegetation to evaporate water drawn from the soil. The water otherwise used by vegetation adds to the moisture in the deeper layers of the soil profile, and some passes beyond the root zone to eventually become streamflow. So an immediate effect is usually seen in headwater streams after a forest fire: the baseflow increases, sometimes several-fold, even before any rain occurs. This has been observed after wildfires and experimental burns in catchments near Eden NSW and in the Brindabella Range ACT.
In the period shortly after fire, and before the catchment's soil water is replenished by rainfall, surface soils can shed water rapidly. This is due to altered soil surface characteristics (described later), and often results in increased runoff during the first rains, especially during thunderstorms.
As long as evaporation from vegetation is suppressed, more and more water becomes stored in the soil during subsequent rain events. The ability of the soil to act as a storage buffer during rainstorms then diminishes, so more runoff can be expected during storm events. Loss of plant canopy therefore can result in flood peaks several times larger than would occur from the unburnt forest.
In the months after wildfire, catchments become wetter than unburnt forested catchments until sufficient leaf canopy returns. Burnt catchments produce significantly larger dry-weather flows and flood peaks, and continue to do so until vegetation re-establishes to attain total leaf areas similar to the unburnt forest.
|streamflow||View Frequently Asked Questions||View Bibliography|
|fire||View Frequently Asked Questions||View Bibliography|
Longer-Term Changes in Streamflow
If pine forests destroyed in fires are allowed to revert to grass or woodland, then water yield from these areas (comprising baseflows and flood flows) will be higher than before the fires. The increases in flows will be approximately equal to that previously lost by canopy interception, i.e., about 2-3 megalitres per hectare of catchment.
If tree death does not occur, as in many Australian native forest types, recovery of leaf area after fire occurs in 3-5 years. By this time, the understorey is also usually re-established. The canopy stabilises, and the water balance reverts to its pre-fire behaviour.
However, where the forest tree species are killed by fire, recovery of leaf area takes a different course. Natural regeneration of the forest occurs, and the regrowth forest is usually far more vigorous than the mature forest it replaced, due in part to access to the pool of nutrients released by the fire. The result is that the regrowth can develop total leaf areas much larger than in the original mature forest. This can occur at age 5-25 years.
The denser canopies in regrowth intercept more rainfall and transpire more water than from the unburnt forest. There is significantly less "left-over rainfall" to appear as streamflow, so water yield from regrowth forest catchments is less than from mature forests.
The outstanding historical example is from the wildfire in the Melbourne water-supply catchments on "Black Friday" in 1939. In that fire, mature Mountain Ash (Eucalyptus regnans) forests were killed. Over the next 30 years, water yield from local areas of regrowth diminished by up to 600 mm/year, or 6 megalitres/hectare. On a catchment-wide basis, where regrowth occupied some 50% of the area, this represented a reduction in annual streamflow of about 24%. By age 75-100, it is expected that water yield will have recovered to the pre-fire condition.
Only forests that are fire sensitive can be expected to show this behaviour. To our knowledge, the sensitive tree species are wet sclerophyll types belonging to the ash family, such as Mountain Ash (Eucalyptus regnans), Alpine Ash (Eucalyptus delegatensis) and Shining Gum (Eucalyptus nitens). Where forests contain a mixture of forest types, the proportion of a catchment carrying these species will produce a similar proportional reduction in long-term water yield when wildfire occurs.
The south-eastern Australian fires of 2001/2002, 2003/2003 and 2006/2007 have occurred primarily in less sensitive dry sclerophyll forests, where we can expect that most of the tree species can survive- damaged, but not dead. Younger trees and understorey may be killed, but these will quickly be replaced by a mixture of species adapted to the ashbed conditions. Therefore, the major reductions in water yield seen after wildfire in the Mountain Ash forests should be regarded as a worst case scenario. The water balance in most burnt catchments (with less sensitive tree species) should revert to their pre-fire conditions within a few years.
In New South Wales, observations in the Sydney water-supply catchments since 2002 suggest that post-fire water yields increase then decline in the more extensively burnt subcatchments (though that cannot be statistically confirmed as yet). However, it appears that water yield in these NSW catchments is not affected in the same significant ways as in the Victorian catchments discussed above. Long-term annual rainfall patterns through the catchments show a constant declining trend over the past 15 years. Long-term annual water yield patterns through these catchments also show a constant declining trend over the past 15 years. Post-fire annual discharge in Nattai catchment, for example, has not exceeded 50% of the pre-fire mean annual discharge (3-20 GL/year vs mean yield over 1965-2001 of 41 GL/year).