Why empirical data are important in wildfire management

The first moments of a wildfire are full of unknowns. Where exactly is the ignition point located? How is the current fire behaviour? In which direction the fire is spreading? Are there critical points or infrastructures nearby? Is access to the area straightforward? It is then that the search for information begins: images (both terrestrial and aerial), meteorological data (from stations in the territory or from meteorological models), operational cartography, historical wildfire perimeters, etc. 

Until the first resources arrive in the area, the level of uncertainty remains high. However, with the data that have been collected since the start of the wildfire, the data from historical fire episodes and the descriptions of the situation from the firefighting personnel, the history begins to be created to explain what the fire wants to do and what the fire can do, which will allow the wildfire management team (incident commanders, fire analysts,…) to draw the strategy to face the wildfire.

Data are of crucial importance in the resolution of wildfires, but it is vital to bear in mind that these data should serve to reduce uncertainty, not to generate confusion. To this end, it is essential to clearly define which data are useful for effective wildfire management, as the mere accumulation of information does not guarantee enhanced management capabilities.

At the Catalan Fire and Rescue Service (CFRS), we use a variety of data sources to gather information during wildfire suppression. This enables us to understand the situation, how it is developing, and to anticipate potential outcomes based on the suppression efforts in place. 

In the following, we will outline some of these data sources, distinguishing between those collected during the wildfire and those gathered prior to it (mainly during historical fires), but which are also necessary for managing the situation.

Data gathered during the wildfire episode

When an ignition starts at a certain point, some data begin to be gathered:

• Field data, such as images and/or videos (both from the surface and air), fire behaviour description (such as flame length - helping to define the intensity -, rate of spread, presence of torching or spotting, etc.), description of the type of fuel involved, in situ meteorological data, etc. 
• Meteorological data, gathered from the meteorological station network, providing real-time information, and from weather models, then helps to know about the evolution of the weather in the next hours and prevent any changes. However, it is important to consider the spatial resolution of the weather models used, as they may not adequately account for the topographic complexity of the area, resulting in forecasts that are not sufficiently accurate for wildfire management. Other weather-related data are the drought indices, to have an idea about the fuel availability in the area. 
• Atmospheric profile data. During the 2021 fire season, in Catalonia, we began collecting atmospheric data using radiosondes. This allowed us to enhance our understanding of the interaction between fire and the atmosphere during wildfires. At the moment, models can't usually predict lower atmospheric conditions accurately. This is mainly because the resolution isn't good enough to read the topography properly. Additionally, these models have no capacity to predict fire-atmosphere interactions. Considering these limitations, our research focuses on acquiring data from the actual atmospheric vertical profile (from both the environment and within the smoke plume) to enhance our understanding of the situation and contribute to the advancement of wildfire science.
• Geographical data. During the evolution of the wildfire, it is interesting to gather data about the actual position of the fire in the territory. This allows efficient management of the extinction efforts and to identify potential changes, critical points or opportunities. Also, when the wildfire is extinguished, it is important to gather the final perimeter, and its evolution, using hourly perimeters. 

This is some of the data that is important to gather during the wildfire episode, and it is crucial to implement a robust structure for the collection and organisation of this information, as it can be highly beneficial in anticipating and responding to future wildfires.

Historical data

Not only the real-time data can help the management of the wildfire to anticipate its spread and behaviour. Having a decent archive of historical wildfires, with information about the final perimeter, its evolution and main runs, the fire behaviour, the meteorological synoptic situation, spread patterns and how the fire service manages the situation, can help to predict the behaviour and the spread of the actual wildfire. 

In Catalonia, using a large wildfire perimeter database, we defined the Wildfire Types, which remains with the idea that in the same meteorological synoptic situation and a similar topography, the wildfire will follow a similar spread structure, a thing that facilitates the identification of critical points, opportunities, main spread directions, etc. Then, using this information with the actual data gathered in the wildfire allows the wildfire management team to elaborate a more accurate forecast of the evolution of the wildfire, facilitating the strategic decision-making process.