This document was produced under the terms and conditions of Grant Agreement No. 101037419 of the European Commission. It does not necessarily reflect the view of the European Union and in no way anticipates the Commission’s future policy in this area.
Copyright © All rights reserved. This document or any part thereof may not be made public or disclosed, copied or otherwise reproduced or used in any form or by any means, without prior permission in writing from the FIRE-RES Consortium. Neither the FIRE-RES Consortium nor any of its members, their officers, employees or agents shall be liable or responsible, in negligence or otherwise, for any loss, damage or expense whatever sustained by any person as a result of the use, in any manner or form, of any knowledge, information or data contained in this document, or due to any inaccuracy, omission or error therein contained.
All Intellectual Property Rights, know-how and information provided by and/or arising from this document, such as designs, documentation, as well as preparatory material in that regard, is and shall remain the exclusive property of the FIRE-RES Consortium and any of its members or its licensors. Nothing contained in this document shall give, or shall be construed as giving, any right, title, ownership, interest, license or any other right in or to any IP, know-how and information.
The information and views set out in this publication does not necessarily reflect the official opinion of the European Commission. Neither the European Union institutions and bodies nor any person acting on their behalf, may be held responsible for the use which may be made of the information contained therein.
The investigations aimed to provide science-based answers through a combination of experimental work and modeling, focusing on three main aspects: (i) the vulnerability of building joinery (windows and shutters) to hedge fires, (ii) the role of vegetation moisture content in fire ignition and spread, and (iii) the effectiveness of fuel management strategies around buildings. Findings highlighted the need for updated building codes mandating fire-resistant construction elements such as double-glazed windows and shutters. The research also demonstrated that vegetation characteristics significantly influence fire behavior near structures. Therefore, regulations must define appropriate vegetation-to-building distances based on plant species flammability and structure exposure, ensuring sufficient separation to avoid direct flame contact with façades and roofs. This approach provides a scientifically grounded basis for fuel management policies in WUI zones.
Currently, the tool effectively pinpoints the most crucial areas for fuel treatments to protect multiple values across the landscape. The next phase of development will integrate specific parameters for the strategic placement of tactical infrastructure, such as linear firebreaks aligned with pathways to aid firefighting efforts and water pools that serve as refilling points for aerial resources. By incorporating the specific guidelines, laws, and operational factors of each country or organization, the tool will provide enhanced support for targeted decision-making in the preparation phase.
The FIRE-RES Innovation Action (IA) 2.4 titled “Optimizing landscape configuration and fire management policies to minimize expected losses from EWE” provides a comprehensive, data-driven framework to protect the WUI landscape by applying land planning and fuel treatment techniques both at the source of their risk, i.e., forested areas several km away from them, and at the “sink” level, i.e., parts of the WUI or core settlement where the incoming fires are expected to burn with higher probability. Fuel break zones and patches of low fuel availability shall act as natural firebreaks, disrupting fire pathways and mitigating risk. The IA answers to where these fuel break zones and patches of low fuel availability should be allocated on large landscapes, considering their potential effect on the reduction of fire intensity, burn probability and exposure of the WUI, settlements and protected areas. It also answers to what is expected from such investments and what is the best way to allocate limited resources across different jurisdictions to maximize benefits.
Designing networks of areas for fuel treatment to facilitate wildfire suppression requires a structured and participatory approach that combines expert knowledge with spatial analysis. The process begins by defining clear objectives, such as improving the effectiveness and safety of suppression operations, and translating them into measurable criteria. These criteria include factors like fuel type, vegetation structure, accessibility, and proximity to communities, all of which influence fire behavior and firefighting capacity.
Using GIS and management units, critical zones are identified and mapped, providing a spatial framework to assess where treatments could have the greatest impact. To ensure the process reflects both technical and social realities, experts and stakeholders are engaged through surveys and focus groups, helping to establish the relative importance of each criterion and to define how utility should be measured.
The integration of multi-criteria decision analysis (MCDA) and the Analytic Hierarchy Process (AHP) allow for weighting the criteria and classifying management units into priority levels. This results in spatial maps that highlight where treatments would be most effective, particularly in areas that serve as access routes, buffer zones, or strategic corridors to slow down fire spread. These priority maps are then used to design a connected network of treated areas, aimed at providing safe zones that enhance suppression opportunities.
Finally, the proposed network is validated and adjusted with the involvement of stakeholders, ensuring its practical applicability under real-world conditions. This methodology has been applied in the Living Labs of Bulgaria, Catalonia, and Portugal, demonstrating how participatory spatial analysis can guide the creation of fuel treatment networks that strengthen wildfire suppression capacity, reduce potential fire impacts, and improve firefighter safety.
Following an extreme wildfire event (EWE), irreversible soil loss may occur if effective restoration actions are not rapidly implemented. After a wildfire, soil loses its protective cover, becoming highly vulnerable to erosion and degradation, which in turn impedes vegetation recovery. Given the scale and impact of EWEs, it is essential to quickly identify the most vulnerable areas in order to prioritize the allocation of the often limited technical and financial resources.
Soil loss is an irreversible impact on a human scale and entails the loss of most of the services provided by the affected system. After a forest fire, the soil becomes drastically unprotected against erosion by losing the protective forest cover, hindering vegetation recovery. The risk of erosion decreases with the recovery of vegetation, which retains the soil and reduces the impact of raindrops and runoff.
Natural recovery of vegetation after a fire depends mainly on the regenerative strategy of the dominant species. In general, plant communities dominated by resprouting species exhibit high survival rates and rapid regeneration of aerial parts after fire, leading to a fast recovery of vegetation cover. In contrast, the recovery of plant communities dominated by germinating species depends on the effect of fire on the seed banks, the composition of the post-fire understory and the meteorological conditions after the fire. Additionally, fire severity and topography affect both soil erosion hazard and natural recovery potential of the vegetation.
Areas with slow plant recovery and a high risk of soil erosion should be prioritized for restoration. Given the limitations of economic and technical resources, efforts should focus on providing spatially explicit tools to rapidly identify these areas, especially when the affected area is large and structurally complex. As the need for restoration can be very urgent, providing forest managers with cartographic information on vulnerable areas through an accessible tool is essential.
Through this Innovation Action, we developed a methodology to efficiently identify high-priority restoration areas by leveraging existing cartographic layers and expert knowledge through expert panel discussions. The main factors influencing post-fire regeneration were identified, assigned weights and utility functions, and integrated to produce a final priority restoration map, designed to support decision-making following an EWE. This approach was applied in the Living Labs of Catalonia, Portugal, and Canary Islands.
The objective is to help decision-makers choose where, when, and how to apply restoration, fire prevention, or management actions (e.g., harvesting, thinning) to balance fire risk, ecosystem services, and economic returns.
More specifically, the following objectives can be distinguished:
To determine which management scenarios (e.g., thinning, harvesting, fire prevention strategies) deliver the best outcomes across multiple dimensions: reducing fire risk (burn probability, flame length, tree mortality), maintaining or increasing ecosystem services (such as carbon stock, timber yield, non-timber products, scenic beauty), and ensuring economic viability (profit, cash flow, NPV).
To compare scenario outcomes both spatially and temporally, so that trade-offs and long-term trends are clearly visible.
To develop spatially explicit tools (maps, priority zones) that help decision-makers locate areas of heightened vulnerability or opportunity – areas where fire risk is high and ecosystem service loss is likely, or areas where restoration or interventions would yield greatest benefits.
To ensure that criteria for assessing scenarios are guided by local stakeholder values and expert judgment, so that weighing ecosystem services, fire risk, and economic indicators reflects what is most relevant in each region.
Still the bigger picture implies that forest management, is to be considered as part of fuel management, and at the same time fire prevention should be accounted into the planning of forest management.
IA 2.8 has been designed to develop and test advanced methods for supporting forest management under wildfire risk and multiple ecosystem service objectives. The approach integrates spatial optimization techniques with a Pareto frontier framework, enabling the assessment of trade-offs among competing objectives such as timber production, carbon storage, biodiversity, and cultural services, while simultaneously addressing wildfire resilience. The milestone presented here summarizes the processes, methods, and technologies employed to address the IA questions, highlights the recommendations emerging from the Living Lab applications, and outlines the future roadmap for exploitation and broader use.
The Chilean and Portuguese Living Labs exemplified a participatory approach where local stakeholders (residents, municipal bodies) selected the test areas. This ensured locally relevant vegetation and structure types were studied, improving relevance of the findings.
Effective fire risk assessment in the WUI requires data on vegetation type and fuel load, moisture content, local weather, topography, and building characteristics (materials, structural elements). These inputs are necessary to assess fire impact on buildings and to carry out numerical simulations of a fire approaching a WUI.
Fire impact studies on buildings were conducted using the EXPLORII experimental platform. Numerical simulations were performed with the CFD code WFDS, which integrates heat transfer and fluid dynamics to model fire behavior in WUI settings.
WFDS is a mature, validated simulation software. Combined with the EXPLORII experimental platform for joinery testing, the project utilized a robust package of tools bridging empirical and modeling approaches.
The study recommends improving building fire resilience through the use of double glazing, the installation of shutters, and fuel management around structures (e.g., creating defensible space). These actions involve modest costs but can prevent significant damage in the event of a wildfire.
The results are publicly available and emphasize practical, action-oriented recommendations. These recommendations are presented to stakeholders involved in wildfire risk prevention, with the aim of being integrated into public fire prevention policies.
By integrating landscape features and weather data from each living lab, recommendations were developed to protect social spaces while promoting active community participation in decision-making processes.
To simulate wildfires, a fuel map for each living lab was required, together with canopy characteristics, historical weather data, and a “values-at-risk” layer representing any valuable element (infrastructure, protected areas, population, etc.) that can be damaged by fire.
Fire simulations were conducted using the C2F-W simulator, applying the fuel classification specific to each living lab. Subsequently, treatment stands were identified through the downstream protection prioritization model, and the expected post-treatment losses were estimated and presented.
FireAnalyticsToolbox is a QGIS plugin that integrates fire simulation modules that calculate key propagation metrics, such as fire intensity and flame length, along with risk prioritization indicators like Downstream Protection Value (DPV) and Betweenness Centrality. It also includes a knapsack optimization model to allocate fuel treatments, primarily based on WUI proximity and other spatial criteria.
The proposed methodology and developed tools do not require any significant additional budget; instead, they suggest a reallocation of the existing firebreak management budget, leading to a more cost-effective configuration that can protect the WUI or other valuable elements/services identified by stakeholders.
The transfer of the developed models is currently underway to ensure that the Chilean Forestry Service (SERNAFOR, formerly CONAF) adopts the firebreak placement and risk prioritization models within the QGIS plugin developed as part of the project. Similarly, ISAs team, from Portugal, was trained to replicate this methodology by using the QGIS plugin.
The Scenario Planning framework, using ForSys in its core, enables land managers to test different solutions and create fuel management plans based on numerical estimates. Governments can use this tool to prioritize investments and control who gets the investment based on the expected benefits of the proposed plans. For most EU countries, where to treat is decided based on local knowledge, randomness or accessibility. The use of spatial fire simulations in an integrated part of the problem-solving algorithm that helps to target treatments to areas with proved fire problems. The proposed framework by this IA can be integrated into the planning phases of National programs such as the “Antinero” (Greece), “Vulcain” (France), “AGIF National Fire Plan” (Portugal) etc. Land managers can use the framework to create and propose new plans for approval accompanied by maps and attainment graphs for different management objectives. Key stakeholders and end-users include Governments (National and Regional), Forest Owners Associations, Logging Associations, Forest Services, Forestry companies that create forest management plans, and electricity companies that need fire planning.
To run the IA planning tool, a key geospatial vector layer—namely the “stands” layer—is required. These are artificial polygons (e.g., hexagons) covering the entire study area, not traditional forest stands. After defining stands, pre-existing planning areas (e.g., Municipalities, Forest Service Districts) are added based on local fuel treatment governance. Management barriers such as protected zones, steep slopes, roads, and recently burned areas are then applied to define treatment availability. Each stand is further attributed with data including wildfire simulator outputs (exposure and transmission), biomass, commercial timber volume, ignition/burn probabilities, carbon stock, and fire intensity.
IA 2.4 is a landscape planning solution that uses fire simulation outputs to optimize fuel treatment allocation. Unlike current practices, it targets areas where investment would most reduce fire exposure to settlements and protected zones. It leverages ForSys, a free and scalable scenario planning tool developed by the USDA Forest Service, using simple spatial inputs to produce maps and rankings of treatment projects. Output can be used to create proposed project maps (shapefiles) and graphs. It solves the problem of finding which areas can better achieve a management goal related to EWE and settlement protection and prioritize them over others that underperform. It eliminates randomness and supports informed, benefit-maximizing decisions. It provides numerical estimates of what is expected to be achieved and ranks projects based on their performance.
Stochastic wildfire simulations provide a probabilistic view of how fires may spread under various ignition sources, weather conditions and land configurations. By identifying fire ignition zones, transmission corridors, and areas of suppression difficulty, simulations allow for the prioritization of interventions such as fuel breaks, prescribed burns, and infrastructure reinforcement. Stochastic fire simulations contribute to cost-effective and spatially smart treatment planning. They are also valuable for simulating post-treatment and future conditions, allowing planners to test the impact of resilience strategies before implementation. Finally, these tools help build risk literacy among planners and communities, fostering improved local engagement and preparedness.
The ForSys tool is central to the spatial optimization strategy in FIRE-RES, as it enables the creation of data-driven treatment plans tailored to local risk profiles and management goals. By inputting spatial layers such as vegetation type, ignition probability, and treatment cost, stakeholders can model various scenarios that balance fire mitigation with ecological, social, and economic goals. The tool supports multi-year implementation strategies, enabling planners to evaluate trade-offs and sequence fuel treatments effectively. It also allows dynamic adjustment of priorities, such as shifting focus between community protection, ecological restoration, or cost-efficiency. This approach ensures not only maximum risk reduction but also fosters cross-agency collaboration and public understanding through interactive interfaces and clear visual outputs.
Scenario Planning with ForSys organizes stand-scale polygons (1–20 ha) into project areas (500–2,000 ha) to meet ecological and logistical goals. ForSys emphasizes user-friendliness, transparency, and stakeholder accessibility. The development of this Scenario Planning tool was motivated by gaps in decision support tools for rendering the growing number of land condition assessments into optimized projects areas as part of prioritization and planning efforts. The Scenario Planning interface was developed with an emphasis for wide application by non-technical users and employs a spatial optimization algorithm that can be easily explained to stakeholders and compared to opaque mathematical programming approaches. It has been successfully applied across varied scales in the USA and Mediterranean Europe, bridging scenario analysis, landscape planning, and decision support.
Use of the tools and outputs from this IA are designed for the National Forest Services and Governments who want to allocate funding to apply fuel treatments to reduce suppression costs and increase settlement protection and minimize the fire effects in protected areas, who are dissatisfied with the current random or non-justified selection of areas for fuel treatments and plans that do not provide a clear numerical estimate on what can be achieved by a multi-million euro investment for fire prevention. The cost to create the input data with trained personnel, conduct the fire simulation modelling and then run fuel treatment scenarios is minimal (<20k euros) and is a one-time investment. The cost to apply the proposed fuel treatments, especially if the target is to achieve landscape level change in expected wildfire dynamics, is very high for each country, but this cost is not directly related to the planning process conducted in this IA. The cost can range from 100m euro for smaller countries, to 1b or 2b euros for countries like Greece, Spain and Portugal.
Effective communication relies on interactive WebGIS platforms showing project locations, rankings, and benefits—far beyond static maps. Graphs enhance understanding of treatment impacts across time. Public and stakeholder education is vital to build support for measures like prescribed burning. Training, policy advocacy, NGO partnerships, and international collaboration (e.g., with the USDA Forest Service) are key to scaling and institutionalizing the tool’s use. Web-based platforms and regular stakeholder meetings further promote transparency, trust, and co-innovation. Partnerships with other land or forest management agencies, like the USDA Forest Service can help to promote the solution and showcase successful implementation in other countries. Participation in focus groups with relevant stakeholders in each country and collaboration with national/regional Decision Makers (Civil protection, Firefighters School, Forest Services) can aid in communication. IA can also offer expertise and knowledge transfer so that partners can use or expand the proposed solution to their area of interest. Technical support and connections with the teams that developed parts of the IA solution from the USA are also possible. Regular partner/stakeholder meetings for capacity action or future collaboration in projects/application of the methodology can help in initiating co-innovation projects.
Stakeholder engagement was a central component of this study to ensure the relevance and effectiveness of the proposed methodology. A broad group of stakeholders was invited to contribute their knowledge and perspectives through surveys designed to capture preferences on criteria and sub-criteria relevant to wildfire management, as well as the opportunity to suggest additional ones.
This process was followed by a focus group discussion that brought together participants from different professional backgrounds, including policy, management, technical, and operational fields. In this setting, stakeholders refined the sub-criteria, assigned weights, and contributed to defining the parameters used to establish utility-function curves, which translate raw data into standardized priority scores. The diversity of perspectives ensured that both ecological and social priorities were considered, while also grounding the methodology in practical management realities.
To generate spatially explicit maps of priority areas for fuel treatment, the relevant criteria and sub-criteria selected with stakeholder input must first be compiled. The information required should be readily accessible, preferably from publicly available spatial datasets, and allow for efficient processing to support timely decision-making.
The methodology focuses on producing a straightforward tool for preliminary prioritization, using mostly cartographic and tabular data. The selected factors are grouped into key categories, such as the potential influence of fuel and vegetation on fire behaviour, the accessibility and proximity to communities, and the logistical feasibility of suppression operations. Each category is represented by sub-criteria that can be derived from spatial data, including forest structure, fuel type, slope and topography, and land-use attributes.
These data are processed and visualized within a GIS environment. Stakeholders provide additional guidance for defining parameter ranges and thresholds, which are used to translate raw values into standardized priority scores through utility functions.
The model framework provides a structured approach to integrate stakeholder knowledge, spatial data, and decision criteria for prioritizing areas for fuel treatment. It is based on a hierarchical organization of criteria and sub-criteria, which allows the relative importance of different factors to be systematically combined.
Raw spatial and attribute data are translated into standardized priority scores, guided by stakeholder inputs on criteria weights and thresholds. These scores are then aggregated to produce maps highlighting areas where fuel treatment would most effectively support wildfire suppression. The framework is adaptable to different contexts and scales, ensuring consistent and informed prioritization across regions.
The technological framework for this study combines spatial analysis, decision-support, and modelling tools to support the prioritization of areas for fuel treatment. Core functionalities include processing and visualizing spatial data, integrating multiple criteria, and translating stakeholder inputs into priority scores. Examples of tools that can be used in such a framework include GIS platforms for mapping and spatial analysis, fire behavior modelling software, and decision-support systems such as Criterium Decision Plus (CDP) and the Ecosystem Management Decision Support System (EMDS). The methodology allowed the use of different systems to show the results and initiate a discussion with the stakeholders.
The framework allows for the hierarchical organization of criteria and sub-criteria, the weighting and scoring of alternatives, and the generation of spatially explicit priority maps. It also incorporates methods to classify and interpret spatial variability and to estimate fire behavior and potential impacts.
The methodology is designed to be practical and applicable within existing management contexts, and its socioeconomic feasibility is enhanced by the use of readily available spatial and tabular data. The budget required to apply the approach in different regions is relatively modest, primarily covering the costs of organizing participatory activities such as surveys and focus group discussions.
The framework can support more efficient allocation of resources for fuel treatment and wildfire suppression, potentially reducing time and human resources needed for field surveys and planning. The approach is intended to guide decision-making in a cost-effective manner and can be adapted to the available financial and operational capacity of the implementing organizations.
Results and outputs of the study are disseminated through scientific publications in peer-reviewed journals, presentations at conferences, and sharing datasets via public repositories. Additionally, project meetings and participatory workshops with stakeholders are used to communicate findings, discuss priority areas, and gather feedback, ensuring that the methodology and its outputs are accessible and relevant to both practitioners and decision-makers.
Engaging relevant stakeholders is essential for ensuring both the acceptance and effectiveness of the developed methodology. These stakeholders include wildfire and post-fire restoration experts from public administrations as well as researchers. Through their participation in expert panel meetings, they select the most important criteria that ensure post-fire restoration success and assign weights and define utility functions to these criteria. The participatory nature of the approach ensures that the selected criteria address both ecological and social concerns and that the areas most in need of restoration after an EWE are prioritized.
To create a final map of high-priority areas for restoration, the post-fire recovery criteria selected by the expert panel must first be compiled. The selected factors should meet three premises:
The information required is freely available and can be accessed immediately.
The priority for restoration is inversely proportional to the capacity of the ecosystem to recover naturally.
The priority for restoration is defined in the short term.
The developed methodology aims to generate a simple tool for the preliminary identification of priority areas for post-fire restoration. For this reason and, to simplify as much as possible data processing, only cartographic data are used.
Based on stakeholder input, the factors selected to prioritize restoration actions are grouped into two main categories: (1) the potential risk of soil erosion and (2) the natural recovery capacity of vegetation. An additional category may also be considered: (3) proximity to social areas.
To characterize the potential risk of soil erosion criterion, three sub-criteria are used: soil erodibility (through the K factor of the RUSLE, Revised Universal Soil Loss Equation), slope length (LS factor of the RUSLE) and fire severity. To characterize the criterion associated to the natural recovery capacity of the vegetation, the following sub-criteria are used: fire severity, aspect, and resprouting ability of the existing community.
Prior to the meeting with experts, key variables for post-fire recovery were identified and obtained from different sources. During the meeting, each criterion was individually reviewed to assess its relevance, and weights and utility functions were assigned by the experts to each criterion based on its importance.
The methodology relies on publicly available cartographic layers or data that can be easily derived through GIS processing of existing information.
All factors are then integrated according to their weights and utility functions into a final equation to determine restoration priority following an EWE.
After preprocessing the base layers, they were integrated to produce a final map identifying high-priority areas for post-fire restoration. This methodology offers a rapid approach to quickly identify the areas in urgent need of restoration following an EWE. The resulting map allows for informed decision-making and resources prioritization in post-fire landscape restoration, and it is particularly useful after large wildfires, as determining where to focus restoration efforts is more challenging.
Because this methodology draws on existing, freely available cartographic data, the budget required to apply it in other regions is minimal and primarily depends on the availability of relevant information for the area being assessed. Moreover, once validated and approved, the methodology can reduce costs associated with fieldwork for identifying priority restoration areas, saving time and human resources.
Results can be disseminated by publishing the methodology in a peer-reviewed journal and by making the generated datasets openly accessible through a public repository.
In every Living Lab, we should convene a panel of local stakeholders (forest managers, fire experts, ecologists, local community representatives) to help define which ecosystem services and fire risk indicators matter most regionally. These stakeholders provide feedback to set criteria weights and utility functions, ensuring that the prioritization aligns with local values (e.g., scenic beauty, mushroom production, fire safety).
In each Living Lab, effective prioritization of management and restoration actions after fires requires the availability of high-quality spatial data [fd1] – topographic layers, digital elevation models (DEMs), slope and aspect information, as well as layers describing forest structure: mean tree diameter, basal area, canopy cover, and fuel types.
Additionally, climate or weather scenarios should be available, accounting for key factors such as fuel moisture, wind direction and strength, and, where possible, fire severity or satellite-derived indices to help estimate canopy loss. Economic data – costs, timber prices, revenues from ecosystem services, discount rates – should also be accessible to evaluate the financial feasibility of scenarios.
Equally important is the involvement of local experts and stakeholders in selecting criteria: only with their input can it be determined which ecosystem services and fire behavior indicators should be prioritized in a specific region. These data, weights, and utility functions must be clearly documented so that the results are transparent and replicable.
In our Living Labs, we applied an integrated methodology that combines a forest growth simulator, a fire spread simulator, and an optimization module – allowing us to assess how different management scenarios affect forest dynamics, fire risk, and economic and ecosystem indicators.
When building the model, it is important that growth outputs – including key stand structural attributes such as canopy height, canopy bulk density, canopy cover, and fuel load – serve as inputs for the fire module, while climate scenarios and management actions such as thinning, harvesting, and other interventions influence both parts of the model. The optimization component should use multi-criteria approaches or other mathematical optimization methods to balance and compare scenarios in terms of economic benefits, fire protection, and ecosystem service preservation.
Criteria (ecosystem services / fire risk / economic) should be normalized and brought to a common scale, and utility functions formulated to reflect the impact of changes in each indicator on the overall benefit of a scenario for stakeholders. The result should be a priority map that integrates all criteria with their weights, providing a visual representation of where interventions are most urgently needed.
In our project, it is essential that the software and technical workflows support both raster data processing and optimization in a way that is efficient, reproducible, and accessible. To this end, we recommend using GIS tools capable of harmonizing raster layers (alignment, reprojection, resampling) so that all spatial inputs (fuel layers, canopy metrics, terrain) share the same grid, cell size, and extent. For raster manipulation – such as calculating zonal statistics, masking, and raster algebra – we rely on libraries like GDAL, Rasterio, and, which offer efficient routines for raster operations.
On the optimization side, modules like PuLP (for linear and goal programming) can be used to solve allocation problems under multiple criteria. Decision-making frameworks (e.g. MCA, Pareto fronts) can be implemented in Python or R to compare and trade-off between ecological, fire-risk, and economic objectives.
For visualization and dissemination, integration with GIS desktop environments or QGIS plugins helps display results, create maps, and allow local stakeholders to explore scenario comparisons.
One of the most challenging requirements, is to link the evolution of the forest with fuel models or fire risk indicators, so the evolution of the forest under different management schedules can derive also into a fire risk assessment.
Given that both financial and technical resources are often limited, prioritize actions that deliver high return per hectare / per unit area. Use per-hectare metrics (profit, cost, NPV, ecosystem service loss) to determine which scenarios are viable in practice. Focus on interventions that maximize ecosystem services retention alongside economic benefits, rather than maximizing volume or profit alone.
Minimize additional costs by using existing spatial data and avoiding requiring new, expensive field data unless critical. Exclude simulations or model runs that do not produce meaningful fire behavior (e.g., small fires under threshold), to focus decision support where it matters.
An important aspect to be considered for this type of action, is that it involves economic actions over lands that often are privately owned, and in many cases by multiple owners. An optimal solution for a landscape may not be perceived as the optimal decision for an individual owner. While the inclusion of forest management is an obligation if a full fire smart strategy is to be implemented, it will require a very strong cooperation among owners, or heavy subsidizing.
Produce maps of priority zones under different scenarios (fire risk, ecosystem service change, economic return) and share them with local / regional authorities. Include summaries or visualization portals where non-technical stakeholders can see trade-offs (e.g., charts comparing scenarios). Publish or deposit the data, assumptions, model settings, criteria weights so that others can reproduce or adapt our work.
Stakeholder engagement has been a central pillar of IA 2.8. In each Living Lab, stakeholders were invited to provide their opinion on relevant objectives, fuel treatment operational constraints, and related aspects. Their contributions shaped both the definition of management objectives and the parametrization of the models. For instance, Portuguese and Greek stakeholders expressed preferences regarding maximum patch sizes and the total share of landscape under treatment, which were incorporated as formal constraints. This co-development ensured that the model outputs were not only technically robust but also socially relevant.
The data and information requirements for the IA were considerable. Geospatial forest stand data, adjacency relationships, forest growth and yield projections, disturbance probabilities, and economic (cost-benefits) data were integrated into a harmonized framework. This enabled the translation of ecological and economic realities into model parameters.
The model framework itself is built upon a mixed-integer linear programming structure that enforces patch-based connectivity constraints while allowing flexible treatment scheduling. By combining this with a multi-objective optimization procedure, we produced Pareto frontiers that illustrate trade-offs among provisioning and regulatory ecosystem services. These frontiers are not purely academic outputs but serve as visual negotiation tools, helping stakeholders identify where modest sacrifices in one service can yield disproportionate gains in another.
Technologically, the IA deployed a package of tools that includes GIS for spatial data preparation, combinatorial mathematical problem solvers such as CPLEX for optimization, and visualization interfaces for presenting Pareto frontier results. The maturity level of these technologies is relatively high: the optimization and GIS tools are widely used in research and professional practice, but their integration in a joint workflow for real-time trade-off analysis is novel. This positions the IA at a demonstration stage, with significant potential for operational uptake.
Socioeconomic feasibility is a central consideration. While the optimization models can point toward technically optimal strategies, their adoption depends on cost, institutional capacity, and willingness of landowners to modify practices. Preliminary analysis suggests that many efficient solutions involve only moderate deviations from business-as-usual timber production while still providing large gains in wildfire resistance and other services. This implies a relatively favorable cost-benefit balance, though further work on budgetary implications and compensation mechanisms is necessary.
Dissemination and communication have been pursued through presentations at Living Lab meetings, joint workshops, and internal FIRE-RES exchanges. The graphical outputs, particularly the trade-off curves and spatial maps, have been key in conveying complex optimization results in accessible terms. These tools have fostered productive dialogue among stakeholders and improved their confidence in the methodology. So far two scientific articles have been published in reputable journals:
In all Living Labs, residents are advised to remove flammable ornamental vegetation such as cypress, maintain high-moisture plants, and establish a 2-meter vegetation-free zone around buildings. In Chile, residents with single glazing are encouraged to upgrade to double glazing and install shutters to protect windows. In Portugal and the Canary Islands, aluminum shutters are recommended when replacing or renovating existing ones.
Guidelines can be generalized across Mediterranean and temperate WUI regions, particularly those with similar building materials and vegetation types. To improve fire resistance of buildings in Wildland-Urban Interfaces (WUI), the following measures are recommended:
Windows: Replace single glazing with double-glazed windows.
Shutters: Install shutters to protect windows from heat and breakage. Aluminum shutters are strongly preferred over wood or PVC, which are combustible.
Roofs and Gutters: Use non-combustible materials for gutters and roof overhangs, and regularly clean gutters to prevent ignition from debris.
Vegetation Distance: Maintain at least 5 m distance between buildings and vegetation taller than 1 m; create a 2 m fuel-free zone around structures to prevent flame contact with façades.
Plant Selection: Avoid highly flammable species like conifers (e.g. cypress) and favor low-combustibility plants with high moisture retention.
Vegetation Moisture: Keep vegetation well-watered to maintain foliar moisture above 55%, ideally above 85%, to prevent ignition.
Tree Spacing: Ensure adequate spacing between trees in managed areas—10 m spacing is effective for eucalyptus according to simulation results.
The recommendations were made for each living lab, indicating the zones that should be prioritized for treatment allocation and according to each fuel and risk map.
The effectiveness and utility of this methodology are highly dependent on the quality of the data representing the landscape and weather conditions of each application area. Given updated data, the following are required: first, the fuel model must be one of the following: Scott and Burgan, Kitral, or Canadian Forest Fire Behavior Prediction (FBP); second, a high-performance computer to run the simulations and compute the optimization models in a reasonable time; and third, a clear definition of the relative importance of the values-at-risk that are to be protected using the methodology.
A key recommendation is the strategic transformation of land-uses near and within the WUI to act as barriers or buffers that slow or stop fire spread. This can be achieved by replacing flammable monocultures (e.g., eucalyptus or dense pine plantations) near the WUI with low-flammability land covers, such as deciduous broadleaf species (e.g., Quercus ilex, Castanea sativa). The designating of “fuel transition zones” between wildlands and communities can create multifunctional land strips (e.g., managed forests, agroforestry, or low-density vegetation areas) designed to reduce fire intensity before it reaches the WUI. Another suggestion is the development of «mosaic landscapes», a practice that can create heterogeneity in vegetation types and structures across the landscape to prevent large contiguous fuel beds. Urban planning must incorporate fire simulation outputs to restrict development in high-risk corridors, as identified by this IA.
Regulations requiring defensible space, fire-resistant landscaping, and accessible road networks for emergency response are essential. To protect the WUI and settlements, we propose adopting spatially optimized fuel treatment strategies. It is crucial to abandon traditional random or road-proximity-based project allocations, which have significantly lower performance in achieving fire risk reduction. Tools like ForSys can utilize optimization to spatially prioritize fuel treatments in areas with the greatest potential. The proposed 10-Year strategic treatment plans can progressively, but with high attainment of WUI and settlement risk reduction even after the first years of implementation, can reduce risk and build landscape resilience. The application of ForSys has demonstrated that spatial optimization can reduce WUI exposure by up to 50% in initial years. These data-driven plans not only guide effective investment but also inform suppression strategies, evacuation planning, and infrastructure upgrades.
To scale this approach across the EU, access to stochastic wildfire simulators and trained personnel is essential. While models differ, their outputs are generally compatible. External consultants or partnerships with universities can help generate required inputs and run simulations. A web interface offering scenario runs with pre-uploaded datasets could further democratize access. Widespread adoption will depend on clear demonstration of success in pilot regions and overcoming resistance to change from traditional fire management approaches. Collaboration with academic and governmental institutions will be key to providing data, technical expertise, and political support.
Internal recommendations from the Living Labs highlight several key factors for ensuring the effectiveness and reliability of the methodology. First, the quality of the input data is critical, as inaccuracies or gaps in spatial and attribute information can strongly affect the final priority maps and model outputs. Second, it is essential to maintain a clear focus on the study objectives throughout the participatory process. This ensures that stakeholder suggestions and preferences are aligned with the goals of the study and are correctly reflected in the hierarchical model and utility functions.
Finally, the AHP process should be carefully explained to participants, clarifying its purpose and how their input influences the prioritization. Providing this guidance helps improve stakeholder understanding and engagement, ultimately enhancing the consistency of responses and the overall quality and reliability of the model outputs.
The methodology has strong potential for generalization across diverse regions and contexts, provided key requirements are met. These include access to high-quality spatial data, capacity to engage relevant stakeholders, sufficient technical infrastructure, and institutional support for participatory planning. Its flexible design allows adaptation of criteria and weighting to local ecological, social, and management conditions. While regions lacking data or stakeholder capacity may face challenges, the framework is broadly applicable to wildfire management, landscape planning, disaster risk reduction, and multi-actor governance contexts.
The methodology ensures an optimal allocation of the often limited technical and financial resources for restoration actions. To ensure its effective application, two key steps are required: (1) gather relevant cartographic data and (2) ensure stakeholder involvement.
The first step involves collecting freely available cartographic data related to soil erosion risk and the natural recovery capacity of vegetation. The availability of these data will depend on the area being assessed. In the living labs (LLs) where this methodology was applied, the necessary information was obtained from public administrations – and included Digital Elevation Models (DEM) and habitat maps - and from satellite images – by estimating the difference of the normalized spectral index of burned area (NBR) between pre- and post-fire dNBR.
It is equally important that relevant post-fire restoration experts actively participate in the process by selecting and weighing the important criteria. This participatory approach is essential to accurately prioritize areas for restoration following an EWE.
The combination of expert-driven prioritization and the use of existing cartographic resources creates a scalable and adaptable framework for post-fire restoration planning, particularly following an EWE in regions with limited technical and financial resources. Through scientific outreach and stakeholder engagement, the presented approach lays the groundwork for broader adoption and long-term impact, contributing to more resilient landscapes and evidence-based restoration strategies.
Each LL should build a priority map that shows not only areas with high fire risk or fuel load but also where ecosystem services are likely to be lost or retained under different management scenarios.
For Soriguera and Kassandra, compare per-landscape vs per-hectare outcomes to identify where interventions are most efficient. For example, if Scenario X gives a high total carbon stock but low per-ha profit or ecosystem service, that may influence which scenario to favor in policy.
Adjust scenario design to reflect local preferences: e.g., if local community values scenic beauty or mushrooms highly, ensure these are included in criteria and given meaningful weights.
Recommend scenario(s) that appear to offer a balance: not just maximum profit or maximum biomass, but good ecosystem service retention and manageable fire risk (as Scenario 3 often does in results obtained for Living Lab Soriguera).
The internal lessons from the Living Labs indicate that stakeholder involvement in parameter setting and constraint definition is essential to ensure credibility and uptake of results. In Portugal, involving local actors in determining patch size thresholds directly improved the feasibility and acceptance of the optimization outcomes. In Greece, the infeasibility encountered due to polygon size underscored the importance of tailoring model settings to local data realities. These cases demonstrate that technical rigor alone is insufficient; adaptive engagement and iterative refinement are equally necessary.
The potential for generalization of the IA 2.8 approach is significant. The methodology- combining fuel treatment patch-based optimization, multi-objective analysis, and participatory calibration (of parameters)- can be applied to other fire-prone landscapes across Europe and beyond. While the specific parameters such as patch size limits or target treatment areas may vary, the conceptual framework is transferable. Importantly, the use of Pareto frontiers as negotiation instruments is not tied to any one region and could serve as a common language for balancing trade-offs in diverse policy contexts.
The roadmap focuses on integrating fire resilience measures into public policies and urban planning, particularly through improved vegetation management and building adaptations. Future efforts aim to replicate the Living Lab approaches in other regions and to develop operational tools to support local decision-making.
The proposed roadmap for exploitation of the IA 2.4 methodology outlines a five-year strategy. In Year 1, research and development should be finalized, with findings published. In Years 2 and 3, expansion to additional EU countries, pilot implementations, and a user-friendly web platform are prioritized. By Years 4 and 5, integration into national policy should be pursued, with treatment plans established at regional and national levels. Beyond Year 5, the goal is EU-wide standardization. Benefits include a projected 15% reduction in burned area and a 10% decrease in buildings impacted by wildfire, depending on government investment. These strategies support efficient use of mitigation funding, increased resilience of landscapes, transparent planning, and improved public engagement. By embedding fire resilience into spatial planning and governance, this methodology supports a transition from reactive firefighting to proactive risk management and long-term adaptation.
In the short term, the resulting spatial outputs support decision-making for local and regional authorities, guiding where to implement fuel management to enhance wildfire suppression and reduce risks. In the long term, the framework contributes to maintaining ecological and social functions by promoting resilient landscapes and informed land management.
Future development includes refining participatory processes, integrating additional data sources and emerging technologies, and developing user-friendly platforms for wider accessibility. There are also opportunities for strategic exploitation through consultancy, training, and decision-support tools, as well as potential adoption across different regions and governance contexts.
Another aspect that is being considered it to use the strategic priority results that this type of plan provides with a more tactical planning. It means that the MCDA process and results can be used into an optimization process to allocate specific actions, or management models on in order to maximize the potential fire mitigation effect, while keeping economical or technical constrains under control.
This Innovation Action has provided a streamlined methodology for identifying priority restoration areas, optimizing both financial and technical resource allocation. The resulting map is intended to support and accelerate decision-making for local and regional administrations as well as for forest and land management entities. In the short term, it helps guide where to place restoration efforts and prevent soil loss. In the long-term, it helps preserve the ecological and social functions of the system by ensuring successful restoration.
In future Living Labs, it is important to maintain a sufficiently long simulation horizon with regular time-step intervals to ensure comparability of outcomes across regions. Maps and priority zones should be validated using ground data and reviewed by stakeholders. Over time, criteria weights may need to be adapted to reflect changes in climate, fire regimes, or stakeholder preferences. Restoration interventions could be piloted first in priority zones to assess their effectiveness and costs before being scaled up. Additionally, providing capacity building, including training and user guides, will help ensure that the methodologies can be effectively applied by local forest agencies.
Long term forest management should not be focused on fire mitigation but include it as an additional goal. We will need to study how to include other more fire-oriented landscape actions into forest or landscape management. While fire mitigation can be considered as a quasi homogeneous problems, needing some adaptation for local conditions, integrating fire objectives into a highly diverse sector as the forest-rural ones will be still a challenge in the future.
Looking ahead, the next steps include completing the composite analyses of subdivided problems and ensuring that results are consistent across sub-regions without spatial connectivity. Further refinements will focus on improving computational efficiency, integrating more detailed socioeconomic indicators, and extending the modeling framework to include dynamic wildfire behavior. Engagement with stakeholders will shift from parameter elicitation toward testing actual decision scenarios, where forest managers can “set targets” along the Pareto frontier and immediately see the spatial management plan implied by that choice.
For exploitation, the vision is to consolidate the workflow into a decision-support platform that integrates the optimization back-end with user-friendly interfaces for data input and visualization. This platform could be deployed by forest agencies, associations, or consulting firms to support multi-objective forest planning under wildfire risk. Beyond direct operational use, the methodology has value in policy dialogue, as it quantifies trade-offs that are often debated qualitatively. From a business perspective, packaging the toolset into a service model, supported by training and technical assistance, could open pathways for sustained application beyond the FIRE-RES project.
In sum, IA 2.8 has reached a critical milestone by demonstrating the feasibility of integrating spatial optimization, Pareto frontier analysis, and stakeholder engagement into a coherent decision-support approach. The next phase will focus on scaling up, refining socioeconomic integration, and ensuring practical exploitation in both local and broader contexts.
