Fires

Geographic Information System (GIS) tool for territorial analysis for documentary cinema on forest fires

The summer of 2025 left a cartography of disaster across Spanish territory. Thousands of hectares of forest burned in a few weeks, transforming the landscape of dozens of autonomous communities.

FUEGO, the film by Elias León Siminiani, takes that scorched territory as raw narrative and research material. To articulate this research, Artefacto developed a geospatial analysis tool capable of crossing the real extent of the fire with the forest composition of the soil, asking which species burned, where, and in what proportion. The result reveals the most affected species in each community and across the entire peninsular geography.

The system integrates three institutional and satellite data sources that respond to different logics. EFFIS, the European Forest Fire Information System of the European Commission, provides the official perimeters of the burned areas, mapped through Sentinel-2 image analysis after the fires are extinguished. NASA FIRMS, NASA’s active fire detection service, works with images from MODIS and VIIRS satellites and records active hotspots, with a temporal resolution of hours and a spatial resolution of 375 meters. These two sources cover different realities of the same phenomenon: one maps the perimeter consolidated after the fire, the other records the thermal signal in near-real time. The Forest Map of Spain from MITECO (2011), known as MFE, completes the analysis with detailed mapping of the composition of Spanish forests, classifying the soil based on the dominant tree species. The processing crosses these layers through geospatial intersection in Python, generating five output GeoJSON layers:

1. The EFFIS layer with European fire polygons. 2. The NASA layer with grouped hotspots. 3. The MFE layer with forest composition. 4. The intersection layer where fire and forest overlap, and 5. A combined layer that unifies both detection sources. Based on these layers, statistics are calculated for total hectares burned, hectares with identified species, and unidentified hectares, thus differentiating areas covered by the forest map from areas where species data does not exist.

The technical challenge of mapping forest fires is not only one of resolution or satellite coverage. The MFE was designed as an instrument for forest management and is updated discontinuously, which means that in many territories its information is years old. The species it records are the dominant ones at the time of the survey, but the forest changes. Recent reforestation, regeneration after previous fires, or land-use transformations may not be reflected. To this is added that EFFIS perimeters are generated from spectral difference thresholds between pre- and post-fire images, a method that may underestimate low-intensity zones or overestimate partially affected areas. NASA FIRMS, for its part, detects active heat and not necessarily spreading fire, which introduces another source of imprecision when seeking to trace definitive limits. The combination of both sources allows for a more robust approximation, but data uncertainty is part of the analysis.

In this context, cinema can provide something that satellite observation systems do not produce, which is physical presence in the territory—direct testimony of what the pixel does not distinguish. This sensory and narrative dimension fills the gaps left open by the data and proposes a way of knowing the territory that goes beyond statistics.

This tool functions as a research laboratory for the film and as a reusable documentary prototype. Within the production process of FUEGO, it allows the team to identify areas with a high density of loss by species, explore territorial patterns that guide narrative decisions, and build its own visual and cartographic archive. The web-deployed interface allows filtering results by Autonomous Community, selecting the fire source, exploring the map over real polygons, and observing the highest burned concentration of each species. The code is modular and can be adapted to other summers, other territories, and other satellite detection sources. Artefacto publishes it as part of its research on automatic perception systems applied to documentaries, convinced that analysis tools are as much a part of cinema as the camera.

Application (requires authentication): https://fuego.up.railway.app/map