Understand the Index v2025

Understand the Index v2025

Introduction

The Greater Hobart Bushfire Exposure Index is built from two complementary layers:

• Wildfire Severity Index (WSI) — a landscape layer that ranks where severe fire is most likely if a fire occurs, based on fuels, terrain and long-termerm dryness. WSI is calculated as a continuous surface and then published as easy to read classes 1–10.
• Wildfire Exposure Index (WEI) — a building level measure of exposure to nearby fuels, computed from the WSI around each building and also shown on a 1–10 scale.

This asset centred approach is designed for preparedness and mitigation planning rather than real time fire spread prediction.

We integrate five inputs to produce WSI and WEI: satellite imagery, vegetation structure, topography, climate dryness, and buildings. Each is versioned and refreshable, so the Index can be updated on a regular cadence while remaining comparable year to year.

• Imagery (Sentinel 2 Level 2A, 10 m): provides the spectral detail to map vegetation reliably across the region.
• Vegetation structure (10 m): combines spectral indices (NDVI/EVI) with a learned tree probability layer to capture the continuity and density of fuels, aggregated over ~1 ha to reflect how fuels behave, then normalised to [0,1].
• Topography: derived from a DEM to capture slope and short range terrain variation; slope modulates potential rate of spread.
• Climate dryness: a slowly varying long term precipitation proxy, resampled to analysis scale; emphasises fuel related effects over vegetated pixels.
• Buildings: quality controlled footprints anchor the analysis to each structure for WEI.
  
  

Figure 1. Geoneon Wildfire Data Stack — Diagram of the inputs → processing → outputs workflow used to generate WSI and building level WEI for Greater Hobart.

How we build the WSI and WEI

1. Vegetation structure (fuels): We fuse NDVI and EVI with a deep learning tree probability surface to map woody vegetation and its local continuity at 10 m. Fuels are then class weighted and aggregated over a ~1 ha neighbourhood to reflect how continuous fuels drive fire behaviour. The result is a vegetation density layer in [0,1].
2. Topographic modifier: From the DEM we compute slope (affecting potential rate of spread) and a local relief measure (short range terrain variation). These are normalised and combined into a single topography factor in [0,1].
3. Climate dryness: We derive a dryness proxy from long term precipitation (wetter areas score lower; drier areas higher), resampled to 10 m. This provides a stable, interpretable background condition for preparedness products.
4. Integration and classes: Vegetation, topography and dryness are combined (unit sum weights) into a continuous WSI and then discretised to classes 1–10 for consistent mapping and communication; class 0 is reserved for masked/no data areas. For each building, we compute the mean of WSI classes (1–10) within 100 m of the building’s centroid to obtain a raw exposure score, then bin it to classes 1–10. This produces an interpretable, per asset index aligned with wildland–urban interface literature that relates structural exposure to the proximity and continuity of nearby fuels.
   
   

Figure 2. Satellite derived vegetation structure (1-5) underpins the fuels signal that drives WSI. Long term dryness (6) and terrain (7-8) provide modifiers to the vegetation driven severity signal.

Data sources for version GH-2025.00

The datasets presented in Table 1 were used for the version GH-2025.00 production run of the Greater Hobart Wildfire Data Stack (used to generate WSI/WEI for this report).

Table 1. Input datasets used for the version GH-2025.00 production run of the Greater Hobart Wildfire Data Stack (WSI/WEI).

Appropriate use & limitations

• Not a spread or ignition model. WSI/WEI do not simulate where a fire will start or how it will move on a given day; they show where severe fire is most likely if nearby fuels burn.
• Input dependencies. Exposure inherits limitations from building footprints and from the design choice of a 100 m exposure radius (a transparent, conservative default).
• Scope changes. Comparisons with prior years should consider area/scope changes and input refreshes.
  
  

Key Benefits

• Strengthened Mitigation Strategies: Improve overall hazard readiness by refining mitigation plans, reducing potential impacts, and enhancing emergency response efforts.
• Community Engagement & Communication: Encourage risk awareness and foster collective action within communities, ensuring informed dialogue and engagement.
• Sustainable Community Recovery: Support long-term recovery by guiding communities in resilient and sustainable rebuilding efforts after bushfires.
• New Resident Education: Educate new homeowners and renters about their specific risks and preventive measures to ensure safety in bushfire-prone areas.
• Resource Allocation & Prioritisation: Maximise the efficiency of resource deployment by focusing on areas that present the highest exposure and potential risk.
• Refined Risk Assessments: Improve the accuracy and reliability of hazard evaluations by conducting more detailed and precise risk assessments.
  
  

Disclaimer

The Greater Hobart Bushfire Exposure Index (the Index) serves as a mitigation tool, designed to enable enhanced understanding and planning in relation to bushfire risks. It is grounded in methodologies that have been reviewed and validated by subject matter experts specialising in natural hazard risk research, bushfire analysis, and forestry. The application of advanced processing methods has yielded results aligned with established bushfire exposure analyses from academic and scientific literature.

Users should be aware that the Index does not delve into the physical properties of buildings, such as roofing materials, deck constructions, and other structural elements. Hazard impacts in surrounding parcels may induce indirect losses in others, regardless of an individual parcel's exposure profile as portrayed by the Index.

By accessing and utilising the Index, users affirm their understanding and acceptance of these inherent limitations and agree to use this tool responsibly.

While the Index available on this platform is developed with care and aims to provide reliable information, it is imperative that users understand and acknowledge the following:

• Limitation of Liability: The developers, contributors, and administrators of the Index do not accept any liability for any damage, loss, injury, or inconvenience suffered because of using the Index. Users are encouraged to use this information responsibly and consider it as one of many tools for informed decision-making.
• No Warranty: The information provided through the Index is supplied “as is” and without warranties of any kind, either expressed or implied. We do not warrant the accuracy, completeness, or fitness for a particular purpose of the information available through the index.
• Not a Substitute for Professional Advice: The Index is not a substitute for professional advice and should not be relied upon as the sole basis for significant decisions. Users are advised to consult with relevant professionals or authorities when considering actions relating to bushfire preparedness, mitigation, and response.
• Dynamic Nature of Data: The environmental data used to build the index is subject to change due to natural processes, and as such, the index should be considered a dynamic resource that represents the conditions at the time of data collection. Users should be aware that actual conditions may vary, and regular updates and consultation of other resources are recommended.
• Data Interpretation: The Index relies on the interpretation of complex data, and different interpretations could lead to different conclusions. Users are urged to consider this variability when using the index.
• Use at Your Own Risk: By using the Index, users agree to assume all risks related to the use of the information provided, acknowledging the limitations and variability inherent in this kind of data analysis.

By proceeding to use the Index contained herein, users acknowledge and agree to the terms outlined in this disclaimer.