Hydrophobicity (or water repellency) is caused by organic compounds from litter and soil organic matter that coat soil particles. Soils in eucalyptus forest are naturally water repellent when they are dry, even in the absence of fire. Fire makes water repellency more homogenous and persistent by heating litter and soil, and causing more deposition of hydrophobic compounds on soil particles.
Sediment infilling is a slow process and it is likely to take many decades or centuries for valley floors to fill with sediment after a debris flow. Therefore, there is likely to be a period after large debris flows when the catchment is less prone to delivering large sediment loads. This is partially the case in post-orogenic landscape of SE AU where slopes are relatively stable and densely vegetated. In regions with active mountain uplift (e.g. California) the rate of sediment infilling is much higher, and sediment availability is not a limitation for debris flow frequency.
Coir logs are biodegradable and can be left in place. Other erosion control structures such as debris barriers in gullies, debris racks and riser at road crossing are best installed as temporary structures and removed after the risk of large erosion events has subsided.
This is a major constraint. At road crossing, the accumulated sediment can be removed from the catchment using excavator and trucks. In less accessible areas, where debris barriers might be installed, there are limited opportunities to remove sediment. These barriers are designed to hold back coarse sediment (cobbled, boulders) and reduce the impact of these on erosion in downstream reaches. They will fill up over time, but the idea is that they would provide some storage during the 1-2 years after fire when risk of large channel erosion event is highest. Vegetation recovery in SE AU forests is fast so after 2 years, the barrier can be removed allowing the build-up coarse sediment to stabilize through natural regeneration of vegetation. There are major logistical/safety/cost/maintenance issues with this strategy that need to be carefully planned prior to installation.
Fire intensity – or fire severity (which is what we measure after a bushfire) – is a major control of erosion risk. High severity fire remove the forest canopy and woody debris, resulting in smooth hillslopes that produce high overland flow velocity and which are exposed directly to the raindrop impact. Our empirical data on erosion rates and our modeling of erosion indicates that the increase in risk is non-liner with increase in fire severity. The risk remains relatively low for understory fire, then increases markedly when there is crown fire.
Yes, there is large potential for using erosion models to optimise fire regimes so that risk to catchment values (water supply, waterways ,etc) is minimised. Using models, empirical data and risk assessment frameworks to understand how planned burning can be used to preserve values will become increasingly important as the case for fuel management through planned burning, cultural burning, etc is building.
I the context of water supply, the asset is water quality (or turbidity) at the offtake where water is extracted from the reservoir. The residual risk in this setting is therefore determined from the days that turbidity exceeds treatment thresholds. A hydrodynamic model, linking sediment delivery from catchments to suspended sediment at the reservoir offtake allowed us to make that link. Similar approaches can be pursued for other assets, such as ecological communities, species, etc by defining ecological thresholds (e.g. duration and concentration of a constituent). This, of course is a complex task, and involved very detailed modelling of sediment/constituent transport and ecological sensitivity. I think we have the frameworks necessary to achieve this type of risk assessment, but probably not the parameters. More research is needed.
There is anecdotal evidence for increased rainfall intensity due to the hating that occurs on burned (black) landscape. I haven’t seen any peer-reviewed works in this yet/ The large wildfire from 19/20 fires season would be the right setting to detect such an effect.
The coir logs (or contour felled logs which are used widely in the US for post-fire erosion control) reduce erosion by providing surface roughness which reduces peak flows and probability of channel initiation at the base of hillslopes. Coir logs also provide some sediment storage, but in steep landscapes this probably a relatively small volume compared to all the material that is available to move. Research have showed mixed results in terms of effectiveness. The density of coir logs, their spatial configuration in a catchment, the catchment attributes and type of post-fire rainfall will all contribute to some degree on performance.
Application of wood shred or mulch might be a better option for hillslope interventions, but this strategy has not been pursued widely in AU because of concerns with weeds and suppression of vegetation recovery. In the US wood mulch and wood shred, applied from helicopters, is the most commonly used strategy. But his method for erosion control is also delivery mixed results, in same way as coir logs. There is no silver bullet. The main thing is to identify erosion hotspots, so that resources towards erosion control are as focused as possible. Research shows that most of the risk is typically embedded within 10% of the burned area. Doing a little bit everywhere is not going to work. Doing a lot in a few spots might make a difference.
No, we haven’t looked at large sale flood processes (and flood alleviation) following bushfire. There will be an increase in flooding risk in first couple of year after bushfire. How catchment risks would play out with changes to flood management techniques whilst exposure to flood increases is not clear to me. It seems, based on literature coming out, that exposure to flood will increase (both in frequency and magnitude) .So whatever management interventions we come up with, these changes in the threat must be understood and accounted for when evaluating risk.
The cost can be approximated by assuming the water would need to be replaced with water from a desalination plant. For Melbourne Water the cost of replacing water from the Upper Yarra with desalination water is in order of ~100 million AUD. And even with this alternative source of water, there is likely to be shortfall in supply, resulting in other social and economic costs to Melbourne’s population.
I am unsure of exactly what references you are referring to, although I think you might be referring to some recent papers by Rachael Nolan, Tonatitzin Tariz and some other collaborators. Whilst I haven’t read these papers in detail, I understand that they mainly pertain to drier forest environments and woodlands, with more sparse tree cover in environments that are more often (or always) likely to be water limited rather than energy limited. The work by Feikema and others was in Melbourne’s catchments, which are in the most part much wetter and actual evapotranspiration is normally energy limited rather than water limited. Mixed species eucalypt portions of Melbourne’s catchments in Feikema et al.’s work did show a smaller peak reduction in evapotranspiration and faster recovery than for the ash species forest types. I think if a more extensive analysis was done, I think you’d find that the results from Feikema et al. are likely to be relatively similar to more recent work by Nolan, Tarin et al., or at least that the Macaque modelling for mixed species forests could be re-parameterized in the model to be closer to the more recent experimental catchment work.
The slide of percentage changes in short term floods were for peaks, which were the numbers that were quoted in the final report. We did model the full flood hydrograph, to produce projected short term changes in flood volumes and inputs to the hydraulic model but I don’t have the changes in volume for the bushfire scenarios readily to hand, unfortunately. The maps of changes in flood outcomes from the hydraulic model should be available from West Gippsland CMA, along with the Lower Thomson River flood study final report (HARC and Venant Solutions, 2019).
This is a really good question but difficult to answer. Feikema et al. modelled some reduced rainfall scenarios, where they scaled back the mean annual rainfall to represent some circa 2005 climate change projections for Melbourne’s catchments. They found that the relative reductions in water yield with reduced rainfall were slightly smaller, i.e. if the trees had less access to rainfall they then used slightly less water in the model. Seasonal shifts may also be important and I think that Feikema et al. only scaled down the rainfall pattern (which was a repeating pattern for a single “average” year), without adjusting the seasonality of the rainfall. For larger projected changes in temperature, some forest types may become unviable at lower elevations and/or lower mean annual rainfall and this could also cause a shift in the spatial distribution of different forest types. I think further modelling should be conducted to understand how more recent climate change projections would influence the runoff results and also to understand the interactions between bushfires, timber harvesting and climate change.
Groundwater extraction is not explicitly represented in either Macaque or BISY, which were the two models that I referred to in my talk. I suspect that groundwater extraction was either sufficiently small to be ignored in the Macaque modelling work for the Thomson or corrections were made to the streamflow inputs and the Macaque models were then calibrated to monthly total inflows to Thomson Reservoir with the groundwater extractions added back.
I am not aware of any plans to re-run changes in water yield post 2019/2020 fires for catchments in the MDB. However, given the significant changes that are possible in at least some catchments of the MDB post 2019/2020, I think that there should be re-analysis of the water yield impact. In addition, I think that there should be some Monte-Carlo analysis that attempts to understand the range of future water yield impacts that could occur due to future potential bushfires and the interactions with climate change and (where relevant) timber harvesting.
Ash forest senescence is the natural process which is reflected by the gradual increase in water yield with forest age, represented by the right hand side of the “Kuczera curve”. This process is already captured in all of the models that I referred to in my presentation, as the models are initialized with a starting age for the forest, which was the last date when there was a major known disturbance. So if there are no other changes (e.g. no change in mean annual rainfall) and no future bushfires, the models already represent a gradual increase in water yield with time due to Ash forest senescence.
EGW explored these options during the bushfires in consultation with industry experts and East Gippsland CMA. It was determined that due to the nature of the bushfires and the extent of the affected catchments, such options were not feasible. Instead we focused on managing the changed water quality within our operations.
Silt Busters are essentially Lamella Clarifiers. They treat water by removing solids and other types of suspended solids from river water. Suspended solid and particulate removal is typically completed by 3 processes called: coagulation, flocculation and sedimentation. It’s the grouping of these 3 processes that is then called ‘Clarification’. There is plenty of information on the internet that further explains these three steps.
East Gippsland Water works closely with DELWP and EGCMA on matters of bushfire preparedness, response and recovery to ensure risks to drinking water quality are understood and managed.
It is estimated that the East Gippsland Bushfires have cost approximately $2 Million. East Gippsland Water’s annual report for 2019-2020 will likely contain more detail once it is released on our website.