Skip to main content
UCP Knowledge NetworkApplied knowledge for action
RISICO

Not a southern Italy that burns, but a forecast that turns into action

By project FIRE-SCENE staffPublished on

By FIRE-SCENE Pilot 3: the CIMA Research Foundation and Regione Calabria

Wildfires

The summer of 2025 confirmed the growing vulnerability of the Mediterranean to wildfires. Heatwaves, prolonged drought and strong winds have favoured longer and more complex fire seasons across many countries in the basin. In Calabria, one of the most exposed Italian regions, this pressure culminated in a sharp peak between 20 and 27 July 2025: a meteorologically critical phase characterised by high temperatures, very low relative humidity, and sustained winds, which triggered numerous fires over wide areas — mainly agricultural and uncultivated lands, but also forested zones and wildland–urban interface areas. During that week, the majority of the region’s total burned area was recorded.

It is within this context that FIRE-SCENE, a European project coordinated by CIMA Research Foundation, was conceived — to transform scientific knowledge of fire behaviour into operational prevention and response capacity.

Models and operations room: a single integrated loop

In 2025, Calabria became a testing ground for fire-risk forecasting. The Regional Administration added a new daily forecast at the municipal and sub-municipal level, complementing the traditional eight-zone bulletin used for forest fire prevention. Issued by ARPACAL, the new bulletin helps local authorities in the Permanent Unified Operations Room (SOUP) make faster, more targeted decisions to prevent and respond to wildfires.

The scientific core of the system is the modelling chain developed and validated by CIMA Research Foundation, particularly the RISICO Calabria model, whose propagation indices are corrected through satellite observations of vegetation water content (NDWI – Normalised Difference Water Index). This variable is operationally adopted by the Region to classify the daily fire danger.

During the most critical week, the joint interpretation of bulletins and model outputs prevented a typical mistake in extreme phases: after the initial peak, the propensity index appeared to decrease, yet detailed parameters showed that this improvement was almost entirely due to a temporary drop in wind speed, while heat and aridity remained critical.
This evidence, made explicit by the RISICO Calabria modelling chain, prompted the operations rooms to maintain high alert levels and ensured timely, proportionate responses.

From forecasted risk to field response

Fine-scale indications translated into very concrete actions: pre-positioning of ground crews and Fire Operations Directors in the most exposed areas; reinforcement of surveillance for early detection and monitoring, including through drones and patrols; and real-time consultation of the meteorological and climatic drivers that guided local fire propensity, to adjust suppression and mop-up operations and minimise reignition risk.

In parallel, the Region strengthened its governance framework: active agreements with the Forest Police, the Fire Brigade, and volunteer associations expanded the detection and logistical support network, including prevention interventions such as the placement of water tanks and equipment in sensitive areas. Tools and organization worked as interconnected systems.

During those days, several fires revealed how narrow the margin for action could be: around 300 hectares burned in San Giovanni in Fiore, a large front between Roccella and Caulonia, temporary closure of the main road link between the highway and SS106 near Maida, and, on the night of 24 July in Cirò Marina, the precautionary relocation of 245 people due to the threat to a seaside resort. Different events, all managed through the same predictive logic: anticipating where risk was highest and intervening before multiple fires became simultaneously unmanageable.

The analyses carried out by CIMA Research Foundation within FIRE-SCENE confirm the scientific robustness of this modelling chain: the RISICO indices, corrected with vegetation indices derived from Sentinel data, accurately identify the days and municipalities where fires actually occurred, with even higher performance for the most severe events. Forecasting, therefore, did not remain on paper: it provided a tangible advantage for the Calabria Region’s operational planning. The next step — already planned by the Region — is to extend this approach to new end users, such as the agricultural sector, since the most effective prevention arises from information that is clear, localized, and shared.

FIRE-SCENE: an European good practice for Mediterranean resilience

Calabria represents an exemplary case study for FIRE-SCENE: a complex territory, highly exposed to risk yet capable of developing integrated responses. Satellite observations, simulation models, and analyses at the regional and Mediterranean scales make it possible to understand not only the physics of fire, but also the conditions that favour its ignition and spread.

This experience illustrates a European good practice: advanced tools, when embedded in an operational centre able to interpret them and in a coordinated territorial network, can transform a wave of danger into a manageable crisis. Under identical meteorological conditions, the combination of fine-scale forecasting, early detection, and cooperation with local forces and volunteers reduced response times and mitigated wildfire impacts, while increasing the effectiveness of suppression operations — ensuring, in particular, the complete containment and extinguishment of active events.

The synergy between scientific research, technology, and institutional collaboration is thus charting a new path toward coexisting with a phenomenon that can be understood and governed. In the language of FIRE-SCENE, every wildfire becomes a scene of knowledge — a meeting point between data, science, and territory — from which emerges a deeper and more resilient vision of a changing Mediterranean.