A decision-making tool comes into play when we need to consider multiple criteria. The aim here is to assess the importance of different criteria and bring them together into a cohesive dataset. This approach is employed to analyse the various elements affecting fire risk, and subsequently, to rank areas based on their composite risk profile.

This assessment tool creates a composite fire risk layer for a designated region by merging multiple hazard layers. These layers contain information on fire danger indices, forest fire damage, and extreme heat hazards. Each hazard layer is categorised and merged into one risk image, offering valuable insights into the overall fire risk within the chosen region.

The Fire Weather Index (FWI) is a globally recognised tool for estimating fire danger based on meteorological conditions. It includes various components that consider the effects of fuel moisture and wind on fire behaviour and spread. Higher FWI values indicate more favorable conditions for triggering a wildfire. In 2007, the European Forest Fire Information System (EFFIS) adopted the Canadian Forest Fire Weather Index System to assess fire danger levels across Europe.

EFFIS provides two indicators: ranking and anomaly, which display the local and temporal variability of FWI as compared to a historical series of about 30 years. These indicators are accessible in the fire danger section's pull-down menu.

The fire danger forecast module of EFFIS employs numerical weather forecasts from two deterministic models: ECMWF and MeteoFrance. The FWI is computed from both the ECMWF model (8 km) and the MeteoFrance model (10 km), offering forecasts for up to 9 days and 3 days, respectively.

The FWI is divided into five levels: low, medium, high, very high (above 38), and extreme (above 50). In June 2021, a "Very Extreme" Fire Danger Class was introduced for regions with FWI values above 70, providing extra information on fire danger in Mediterranean areas.

Historical data on forest fire loss is vital for managing fire risk. This data helps to understand the build-up of flammable materials and the vulnerability of forested regions. By evaluating forest fire loss over the past 15 years, we can identify areas susceptible to recurring fires. This assessment includes evaluating the accumulation of flammable materials such as dead vegetation and fallen branches, which can fuel wildfires and increase ignition and spread risk.

The Forest Fire method analyses an image to create risk scores linked to the distance from previous forest loss areas. It measures the distance between areas of interest and other parts using a special calculation method. It needs an image and a target area as inputs. The relevant part of the image is then extracted and limited to the target area. A mask is created to identify areas of interest and other parts. The method measures the distance between these two types of areas, with a maximum distance limit. It then evaluates how close the other parts are to the areas of interest based on set distance limits. Finally, it sorts these closeness scores and produces an image showing risk scores related to the distance to the areas of interest.

Tyukavina, A., Potapov, P., Hansen, M.C., Pickens, A., Stehman, S., Turubanova, S., Parker, D., Zalles, V., Lima, A., Kommareddy, I., Song, X-P, Wang, L. and Harris, N. (2022) Global trends of forest loss due to fire, 2001-2019. Frontiers in Remote Sensing

Analysing extreme heat events is crucial for fire risk assessment. It allows us to pinpoint regions prone to high temperatures, which significantly affect fire behaviour and ignition potential. Extreme heat exacerbates wildfire conditions by drying out vegetation and fuelling fire expansion. Incorporating this data into fire risk analysis can identify areas with an increased risk of wildfires.

The extreme heat hazard dataset is a global layer of data for extreme heat hazards. It is classified based on a widely accepted heat stress indicator, the Wet Bulb Globe Temperature (WBGT, in °C), specifically the daily maximum WBGT.

The Wet Bulb Globe Temperature (WBGT) is significant not only for human health but also for various projects and sectors, including infrastructure. Heat stress affects personnel and stakeholders, influencing the design of buildings and infrastructure. Scientific literature often employs WBGT in heat stress studies, applying thresholds of 28°C and 32°C to categorize heat stress risk. The risk levels are designated as slight/low (<28°C), moderate/high (28-32°C), and severe/very high (>32°C).

The process of applying thresholding to this section for risk map evaluation is notably straightforward, given that each factor involved in this analysis is accompanied by specific value limits. This simplicity aids in ensuring that the methodology is both transparent and replicable.

Turning our attention to the FWI approach, we embarked on a comprehensive process of counting monthly extreme events derived from daily FWI data values. This was conducted over an extensive period of 25 years. The objective was to capture every instance where the FWI index surpassed the significant threshold of 38. In doing so, we were able to identify a pattern, suggesting that the particular region under study exhibits a historical susceptibility to spiking FWI values.

This provides invaluable insights into the fire risk profile of the area, allowing for more informed decision-making when it comes to fire management strategies. It is important to note that the frequency of these peak events is just as critical as their occurrence. Therefore, based on the cumulative count of these events, if an area had more than 20 such peak incidents within the 25-year period, it was immediately flagged as a high-risk region.

This strategic identification of high-risk regions serves as a crucial step towards proactive fire risk management, enabling authorities to prioritise resources and efforts in areas that need them the most.

In relation to forest fire losses, we employ the cumulative distance cost approach to identify areas near historical fire loss incidents. Regions close to past forest fire events receive a high risk score, while those further away receive a lower score. This is due to the fact that areas with a history of forest fire losses are likely to contain flammable materials, making forested regions highly vulnerable.

Lastly, in terms of extreme heat hazards, the thresholds are predefined in the provided dataset. Therefore, the values are masked based on those thresholds.

The final output is a visual representation that colour codes the regions based on their calculated risk. The regions are classified as:

Fire risk areas New Version - green (low), yellow (medium), red (high)