The 2019/20 bushfires shocked us all. Their extent, duration, severity and impacts to date have been unprecedented, directly impacting much of Australia’s eastern seaboard landscapes and touching a staggering 80% of the Australian population.
While the impact of fire to catchments presents many challenges, we believe our sector has potential to control, mitigate and minimise these impacts through well targeted management interventions.
Through an internally funded initiative, Alluvium has mapped fire severities and modelled post-fire erosion risk for 6 million hectares of burned areas in temperate forests of eastern Australia. The work draws on the most recent research on bushfire hydrology in Australia and has been carried out in collaboration with Forest Hydrology Group at The University of Melbourne. The outputs from the project includes maps of fire severity, pre-fire erosion rates and post-fire erosion rates at a 1 x 1 km resolution. These maps are provided as high-resolution images in the links below.
We hope that this work will enable catchment, water and asset managers to gain a deeper appreciation of the impacts of the 2019/20 fires in their jurisdictions, and directly inform risk assessment, management interventions and monitoring programs.
Our plans are to continue updating these products as additional post-fire Landsat imagery is released and providing more in-depth factorial analysis of the impacts of erosion on water supplies, key ecosystems and communities. Please contact us if you’d like to discuss the applications or implications for your work.
Details and limitations of the work
Post-bushfire erosion risk is modelled using the Revised Universal Soil Loss Equation (RUSLE). For undisturbed conditions, estimates of erosion rates from RUSLE (and the factors which underpin these estimates) are available from CSIRO at ~1km resolution (Rossel et al, 2016).
In this project we make adjustment to two factors in RUSLE that are strongly affected by bushfire. These factors are related vegetation (C factor) cover and soil erodibility (K factor). The methods for incorporated fire-effects were developed by forests hydrologists at The University of Melbourne from empirical data and literature review on bushfire hydrology (Sheridan et al, 2019). We note that the scientific understanding of fire impacts has not yet progressed sufficiently to enable quantitative assessments of post-fire erosion that fully address the variability and complexity of interactions between the many factors influencing post-fire hydrologic and sediment regimes across the varied landscapes in Australia.
However, the structure of the algorithm and the parameters used in this project reflect the latest available science from Australia and internationally (Van der Sant et al., 2018; Blake et al., 2020; Langhans et al., 2016; Noske et al., 2016; Nyman et al., 2019; Sheridan et al., 2016; Moody et al., 2015; Nyman et al., 2013). The drivers of variability in post-fire erosion, represented by the model, are fire severity (mapped with Landsat 8 imagery; Key & Benson, 2005) and aridity index, mapped using rainfall and potential evaporation grid from the Bureau of Meteorology.
Results from the erosion modelling are broadly consistent with available observations, with increases in erosion (in year 1 after bushfire) ranging from 1 (no change) to a maximum of about 100 (two order of magnitude increase) depending on fire severity and background soil erodibility. However, we note that the estimates of annual erosion rates are generated for areas where we have no data on post-fire erosion. Therefore, the maps should be used as a qualitative indicator of erosion risk and regional assessment of soil loss and potential changes to sediment regimes. The metrics are not suitable as quantitative input into detailed analyses and modelling of catchment processes and risks. They should, however, help inform where more intensive modelling and data-driven risk assessment should be prioritised.
RUSLE is not a physical-based model. It doesn’t predict erosion event or catchment-scale sediment yields. It’s a tool that can provide an indication of the spatial and temporal impacts on soil loss from burned areas. RUSLE has advantages over other empirical or physical models in its simplicity, requiring less direct field data and enabling more rapid post-fire estimations of annual soil erosion. However, being a relatively simple model means that it doesn’t consider processes that are important for linking hillslope processes with sediment transport at catchment scale. Also, the erosion metrics describe the average annual response. The actual erosion response in any given post-bushfire period is highly dependent on intensity and timing of rainfall events.
Our analysis across the entire eastern seaboard has been conducted on a 1 x 1 km grid, commensurate with available date from CSIRO (Rosso et al, 2016) and the level of field calibration and validation. More detailed investigations with targeted validation would enable this approach to be extended to much finer resolutions e.g. the supporting remote sensing imagery adopts a 25 x 25 m grid.
Anecdotal information we’ve received so far from catchment managers has reaffirmed our analysis, with areas of high expected post-fire sediment yield in north-east Victoria having already experienced highly sediment laden flows during recent rains (e.g. Nariel Creek and the upper Murray River).
Citations
Blake, D., Nyman, P., Nice, H., D’Souza, F., Kavazos, C., & Horwitz, P. (2020). Assessment of post-wildfire erosion risk and effects on water quality in southwestern Australia. International Journal of Wildland Fire.
Key, C. H., & Benson, N. C. (2005). Landscape Assessment: Ground measure of severity, the Composite Burn Index; and Remote sensing of severity, the Normalized Burn Ratio.
Langhans, C., Smith, H. G., Chong, D. M. O., Nyman, P., Lane, P. N. J., & Sheridan, G. J. (2016). A model for assessing water quality risk in catchments prone to wildfire. Journal of Hydrology, 534, 407–426. https://doi.org/http://dx.doi.org/10.1016/j.jhydrol.2015.12.048
Moody, J. A., Ebel, B. A., Nyman, P., Martin, D. A., Stoof, C. R., & McKinley, R. (2015). Relations between soil hydraulic properties and burn severity. International Journal of Wildland Fire, 25(3), 279–293. https://doi.org/http://dx.doi.org/10.1071/WF14062
Noske, P. J., Nyman, P., Lane, P. N. J., & Sheridan, G. J. (2016). Effects of aridity in controlling the magnitude of runoff and erosion after wildfire. Water Resources Research, 52(6), 4338–4357. https://doi.org/10.1002/2015wr017611
Nyman, P., Sheridan, G. J., Moody, J. A., Smith, H. G., Noske, P. J., & Lane, P. N. J. (2013). Sediment availability on burned hillslopes. Journal of Geophysical Research: Earth Surface, 2012JF002664. https://doi.org/10.1002/jgrf.20152
Nyman, P., Box, W. A. C., Stout, J. C., Sheridan, G. J., Keesstra, S. D., Lane, P. N. J., & Langhans, C. (2019). Debris-flow dominated sediment transport through a channel network after wildfire. Earth Surface Processes and Landforms, n/a(n/a). https://doi.org/10.1002/esp.4785
Rossel, RAV., Teng, H., Zhou, S., Behrens, T., Chappell, A., Bui, E (2016): Maps of Australian soil loss by water erosion derived using the RUSLE. v1. CSIRO. Data Collection. https://doi.org/10.4225/08/582cef2dd5966
Sheridan, G., Nyman, P., Langhans, C., Cawson, J., Noske, P. J., Oono, A., et al. (2016). Is aridity a high-order control on the hydro–geomorphic response of burned landscapes? International Journal of Wildland Fire, 25(3), 262–267. https://doi.org/http://dx.doi.org/10.1071/WF14079
Sheridan G, Lane P, Noske P and Nyman P (2019) Post-Fire Soil Loss Adjustment Functions for the Revised Universal Soil Loss Equation (RUSLE). University of Melbourne, Draft Technical Report for the Victorian Department of Environment Land Water and Planning (DELWP), November 2019. pp 8
Van der Sant, R., Nyman, P., Noske, P. J., Langhans, C., Lane, P. N. J., & Sheridan, G. J. (2018). Quantifying relations between surface runoff and aridity after wildfire. Earth Surface Processes and Landforms. https://doi.org/10.1002/esp.4370
