Analysing Wildfire Risk for South Central Wales
Requirement
Following a competitive tender process, gi Perspective were commissioned by Natural Resource Wales, the Welsh Government body responsible for environment and natural resources, to undertake a wildfire risk analysis and GIS layer development. Our remit was to use GIS datasets, including data on historical wildfires, to create a wildfire risk index for a region of South Wales. We worked in collaboration with the ‘Healthy Hillsides’ project team, whose work aims to minimise the impact and severity of the frequent wildfires in the South Wales Valleys through better natural resource management.
Solution
Our detailed plan set out the key project stages, from Engagement and Data Discovery, through Model Conceptualisation and Design, to the Risk Analysis and Refinement. Key risk elements were identified and weighted through a combination of expert stakeholder input, literature review and statistical analysis. A risk map is only as strong as the data inputs, so considerable time was spent reviewing and improving key datasets, notably the vegetation habitats data which describes the potential fuels for wildfires. Meanwhile, a sophisticated model was developed, using raster analysis in QGIS to quantify the risk of Ignitions, wildfire Spread and Impacts for each 25m x 25m cell across the study area. Initial risk scores were reviewed by the project team and other local stakeholders, and the model was revised to ensure that it was capturing the situations of greatest concern for local firefighters.
Result
The assessment identified the 3 key risk factor datasets as historical wildfire ignitions, land cover habitat type and buildings. Lower weightings were given to many other layers including deprivation, slope, countryside access layers and exposed infrastructure. Wildfire risk was quantified across the study area and mapped as raster and vector layers, with PDF maps for stakeholders also highlighting the highest risk areas where interventions would be most valuable. The method used proved an efficient tool for approximating wildfire risk for those without recourse to specialist wildfire modelling software which can model the dynamic spread of fire across the landscape under particular weather conditions. Initial feedback for a few key fire station areas has confirmed that the risk map matches with local expectations and insight about areas of greatest risk.
Images: whitcomberd/stock.adobe.com
