Abstract
Portugal is a country that is often affected by large wildfires, which can be characterised by a high level of destruction of material goods and deaths. Portugal will never forget the fatalities linked to forest fires in recent years; in 2017, more than 100 people died. Vila Real is one of the districts with the highest occurrence of wildfires. This study characterised the risk of fire for the District of Vila Real, based on Geographic Information Systems. This work applies a methodology at a district level that allows better decision making by those responsible for district civil protection, since they have a new decision making support tool. This new tool gives information about the wildfire risk, the priority of surveillance of the most dangerous areas and the travel time of the firefighters in their emergency response area. Most of the district (61.78%) presents a medium, high or very high risk in the context of wildfires. Regarding travel time, most of the territory under study is more than 15 min from the fire brigades. Since 70% of the district’s area is not visible from any lookout post or only from one, the areas in need of enhanced surveillance were determined.
Wildfires can occur naturally (e.g. lightning) or due to human factors (negligence or intent). Human factors represent the majority of ignitions that occur in rural areas in Portugal (ICNF, 2020). Climate change around the world as well as other factors (e.g. poor management of fuel or a lack of labour in rural areas) make wildfires more frequent and severe, such as large wildfires in recent years (Di Virgilio et al., 2019). These large wildfires have been occurring all over the world, mainly in the USA, Australia and the Mediterranean area of Europe, including countries such as Italy, France, Greece and Portugal (Ribeiro et al., 2020a, 2020b; Tedim et al., 2018).
In Portugal, severe wildfires have been witnessed, not only due to the high extension of the burnt area, but mainly due to recorded personal accidents, some of them fatal. Besides the high loss of materials and goods, the loss of human lives has been particularly felt. From 1931 to 2020, 387 deaths associated with wildfires were recorded (CEIF, 2020; Molina-Terrén et al., 2019). Portugal faced a fateful year in 2017, which left 116 fatalities, hundreds injured and bereaved relatives in immense pain forever. This tragedy should be remembered and prevented. Portugal will never forget this fatality, which was mourned nationwide and wounded the remaining relatives, forever marked in the bodies of some and in the memory of all. Great fires are strongly marked in the memory of the populations, due to the extreme and devastating spread of such fires, causing loss of properties and lives (Viegas et al., 2017, 2019).
The knowledge of fire behaviour is an essential tool to fighting it and increase the safety of firefighters and of the populations living in the areas of greater risk of wildfire. According to Viegas et al. (2011), fire behaviour can be classified as normal or extreme. Accidents related to wildfires are often due to insufficient knowledge about fire and its behaviour, especially regarding extreme fire behaviour (Viegas, 2006). Most behaviour models are based on the assumption that the properties of fire propagation are quasi-static and can be determined considering three essential factors: topography, vegetation and meteorology, which are known collectively as ‘the fire behaviour triangle’ (Brown and Davis, 1973). However, Viegas (2006) proposes, as an alternative to this classic concept, the ‘the fire behaviour square’ concept, adding a fourth factor: time.
The evolution of computational power has allowed the appearance of fire modelling systems and developed software through scientific investigation (Altintas et al., 2015; Finney, 2004; Kalabokidis et al., 2016; McFerren and Frost, 2009; Sharples et al., 2012). The modelling of extreme fire phenomena allows a better understanding, through empirical, semi-empirical or physical models (Sullivan 2009a, 2009b).
Several wildfire modelling works have been published scientifically to contribute to a deeper understanding of fire behaviour (Lopes et al., 2019; Morvan et al., 2009; Santoni et al., 2000; Weber, 1991). All this computational power is essential to improve the characterisation of wildfire risk. Knowing the risk, it can be managed better and the planning of resources and means can be better ensured as well.
Wildfire risk mapping has a fundamental role and can and should be used as a prevention and planning tool. Risk mapping defines the best location of resources and infrastructures for the defence against forest fires and the optimisation of the measures to the forest surveillance, taking into account the priority areas that are identified as having the highest fire risk. The calculation of fire risk involves variables of biophysical origin (land occupation, location, and type of combustible material), physiographic variables (slopes and exposures) and social variables (demographic density and path density).
Through the collection and treatment of official data, using geographic information systems, the intention is to characterise the risk of wildfire in the District of Vila Real by identifying areas that are prone to the spread of fires with great speed and intensity that will translate into risk areas both for the general population and for the firefighters, who fight these fires with greater attention to ensure their safety.
The District of Vila Real is located in the Northern Region of Portugal and is divided into the subregions of Douro, Alto Tâmega and Ave. This district is bounded to the north by Spain, to the east by the District of Bragança, to the south by the District of Viseu and the west by the districts of Porto and Braga.
The District of Vila Real has an area of 4,307 km2 and is the 11th largest Portuguese district. The district is divided into 14 municipalities: Alijó, Boticas, Chaves, Mesão Frio, Mondim de Basto, Montalegre, Murça, Peso da Régua, Ribeira de Pena, Sabrosa, Santa Marta de Penaguião, Valpaços, Vila Pouca de Aguiar and Vila Real (district headquarters), as seen in Figure 1.
Figure 1 Geographical framework of the District of Vila Real
The knowledge of a region’s climate is essential for planning and managing socio-economic activities and is also essential for mitigating the consequences of risks.
The Douro Region, where the District of Vila Real is located, has been affected by extreme weather events due to climate change (Almendra et al., 2017). The north of Portugal has been characterised by extreme events that have longer drought period during summer and increased intense rainfall in winter (Nunes et al., 2019).
Extremely high summer temperatures lead to droughts and dry out vegetation making it easier for an anthropologic cause to ignite a fire (IPCC, 2014). Rainfall has a significant effect on forest fire danger (Giannakopoulos et al., 2009). In the rainy season, water promotes the rapid growth of fine fuels (Živanović et al., 2020). Depending on the period of the year, it can contribute to increased fine fuel growth and therefore increase fire danger during the fire season (Viegas and Viegas, 1994). There is a direct relationship between the climate and the problem of wildfires, with the climatic elements of greatest interest being the average air temperature, precipitation, the level of sunshine and the relative humidity of the air (Almeida et al., 1995).
The summers in the District of Vila Real are extremely hot and dry, and winters are extremely rainy. These combined variables increase the wildfire risk, as described above (IPMA and AEM, 2000).
The District of Vila Real is very mountainous, reaching the highest altitude in the Serra do Larouco, Montalegre Municipality, at 1525 m. The District of Vila Real is crossed by the initial course of the Cávado River and its tributary Rabagão and by the rivers of Tâmega, Corgo, Pinhão and Tua, tributaries of the Douro River. Large hydroelectric projects have been developed on the Douro River, including the Bagaúste Dam, located between the districts of Viseu and Vila Real.
A fire at the base of a slope or canyon spreads slowly initially and with a shape close to an ellipse (Alexander, 1985; Finney, 2004; Viegas, 2002). Over time, the fire front intensifies due to the heat it receives from the flames of the flanks, thus increasing the height of the flame of the fire head and, consequently, its rate of spread (Butler et al., 2007). The upper flame height needs more oxygen for combustion, more air entering the front of the fire, leading to a further increase in the rate of spread and so on. The increase in fire spread and intensity is a consequence of the interaction between the flame and the ambient air. It is, then, a self-feeding process that is repeated and intensified over time and that leads to an increase in the rate of spread (Finney et al., 2015; Liu et al., 2021). This phenomenon is especially observed in canyons, although it can occur on slopes with a slope greater than 30° (Dold and Zinoviev, 2009; Dupuy, 1995; Dupuy and Maréchal, 2011; Dupuy et al., 2011; Rodrigues et al., 2019; Silvani et al., 2012; Viegas, 2004; Viegas and Pita, 2004; Viegas et al., 2011). The fire can then, in a short time, cover a large area of land, with a capacity for destruction that, in most cases, is practically impossible to avoid (Liu et al., 2021; Rodrigues et al., 2021; Viegas and Simeoni, 2011). This phenomenon is a consequence of the convection produced by the fire itself, which translates into high-intensity winds and, naturally, a high rate of spread. This process expresses the dynamic behaviour of the fire, which is given by the continuous increase in the speed of propagation over time (Rodrigues et al., 2019; Viegas, 2006; Xie et al., 2017, 2020).
The early detection of a wildfire is essential to control the fire spread and, consequently, the success of the combat. Therefore, it is important to study the visibility of certain areas of the District of Vila Real from the national network’s lookouts, to find covered areas and also areas out of visual reach, this being a criterion that can contribute to the potential risk of fire (Alkhatib, 2014). A large part of the district is not visible from any lookout point or only from one. It is also important to mention here that the construction of the visibility chart for the lookout posts was based on an observer 1.70 m high (added to the platform height), a target height of 20 m and a radius of vision of 15 000 m.
Most of the use of forest soil (67.00%) or the use of land corresponding to shrubs (68.31%), is not visible from any lookout post or just from one. In the set of the two land uses, 67.55% of the corresponding area is not visible by any lookout post or only by one of the lookout posts, with 32.45% being visible by two or more lookout posts. If it is considered that at least one lookout post can give a reference to the location of the fire, although not as accurately as two lookout posts together that allow the crossing of two angles on the map, it can be said that 69.69% of the area of the District of Vila Real, corresponding to forests and shrubs, is covered by one or more lookouts.
The wildfire risk mapping was based on the Technical Guide for the elaboration of Municipal Plans for the Defence of the Forest Against Fires (PMDFCI) from the Institute of Nature and Forest Conservation (ICNF) (ICNF, 2012).
The elaboration of the presented maps was made using an open-source software of geographic information system, QGIS 3.10.5 ‘A Corunã’. In this process of cartographic production, multiple techniques for the analysis of geographic information were used, including the Triangulated Irregular Networks (Tin) analysis, for the development of a digital terrain model for viewing, for example, slopes or exposures; tools of spatial analysis such as ‘Viewshed Analysis’ to calculate the lookouts visibility; and the use of additional plug-ins such as the ‘Road Graph’ for calculating the travel times of the fire departments or only the spatial analysis through raster information.
The features of the risk model are presented in Figure 2.
Figure 2 Risk model adapted from ICNF (2012)
Probability is the possibility of a phenomenon occurring in a given location under certain conditions, indicating the annual probability of a rural fire occurring. The susceptibility of a territory reflects the conditions that this area presents for the occurrence and the potential for damage, indicating the potential for fire severity in that area. The degree of danger is determined by multiplying the probability by the susceptibility (ICNF, 2012). According to Varnes (1984), danger is ‘the probability of the occurrence, within a certain time interval and a certain area, of a potentially harmful phenomenon’.
Varnes (1984) defines vulnerability as ‘the degree of loss of an element or set of elements resulting from the occurrence of a natural phenomenon of a given magnitude’. Suarez (2002) presents this definition for vulnerability: ‘the ability of a system to be damaged by some disturbance. It is the function of the probability of occurrence and its magnitude, as well as the capacity of the system to absorb and recover from such disturbance’. Vulnerability defines the degree of loss to which an element at risk (e.g. population, properties and economic activities) is subjected, designating its capacity to resist the phenomenon and to recover after it. Vulnerability is expressed on a scale from 0 to 1, where 0 means that the element is impervious to the phenomenon, with no damage occurring, and 1 means that the element is destructible by phenomenon (ICNF, 2012).
The economic value makes it possible to quantify the investment necessary to recover an element, depending on its vulnerability, after destruction or loss of performance due to exposure to a harmful phenomenon. The extent of potential damage is determined by multiplying economic value by its vulnerability. An element that has a high economic value and is invulnerable will have zero potential damage as it will not be affected by the phenomenon. However, the potential damage will be all the greater when the vulnerability is close to 1 and its economic value is high (ICNF, 2012).
The level of risk is determined by multiplying danger by potential damage. Bachmann and Allgöwer (1998) mention that the wildfire risk is ‘the probability that a wildfire will occur in a specific location, under certain circumstances, and its expected consequences, characterised by the impacts on the affected objects’. Thus, it can be said that risk can only be spoken of in terms of probability, susceptibility, vulnerability and economic value.
Vila Real is one of the country’s districts most affected by wildfires. Between 1975 and 2018, wildfires in the district consumed 472 152 ha of forest and shrub, with 51.34% of that area (242 403 ha) consumed between 2000 and 2018. There are areas with recurrence of wildfires, especially in the municipalities of Alijó, Boticas, Mondim de Basto, Montalegre, Ribeira de Pena, Valpaços, Vila Pouca de Aguiar and Vila Real. The Municipality of Montalegre is the one with the largest burnt area, and from 1975 to 2018, its total burnt area in this district is 21.48%. In the last decade, it is also the Municipality of Montalegre that predominates regarding burnt area, with 23.94% of the total burnt area in the entire district between 2010 and 2018. The Municipality of Montalegre represents 18.72% of the district of Vila Real.
The land use that predominates in the District of Vila Real, occupying 37.76% of its territory, is forest use. Regarding forest occupation, Maritime pine (Pinus pinaster) predominates, with 23.59% of occupation of forest use. Other species with expression in the district are oaks and broadleaved trees. Holm oak (Quercus rotundifolia), invasive species and umbrella pine (Pinus pinea) are rare species in the district, assuming values very close to 0%.
Two other expressive land uses are shrubs and agricultural areas, which occupy 27.51% and 26.00%, respectively. The predominance of forests is found in the Municipality of Montalegre, although there is a little throughout the district, with less expression in the municipalities of Mesão Frio and Santa Marta de Penaguião. Agricultural areas (26.00%) predominate in the Municipality of Valpaços (5.45%), followed by the municipalities of Chaves (4.47%), Alijó (2.54%) and Montalegre (2.53%).
Topography also has its influence on fire behaviour. The main characteristics of the relief that must be considered are the forms of the relief, the altitude, the presence of water lines, the exposure and the slope of the terrain. The temporary water lines at the joining of two slopes forming embedded valleys, with steep slopes, will induce a completely different behaviour from what would exist in a flat area or on a terrain with a gentle to a moderate slope with the same type of fuel. In the embedded valleys, a sudden increase in the speed of propagation and the intensity of the flames can be observed, which may cause an episode of eruptive behaviour.
In the District of Vila Real, there is a minority of area with plains or plateaus and areas with a slope predominance, reaching slopes greater than 40% in all municipalities. The steep slopes can be more powerful for wildfires because when the fire progresses up the slope, the inclination of the flames is for the fuel not yet burnt, leading to its preheating and an increase in the radiative heat flow and, consequently, at a faster rate of spread.
Assessing the risk of wildfire involves studying human behaviour. Thus, it is important to analyse the proximity to the main roads and the density of agricultural and forest paths, as these are two factors that interfere, in large part, with the danger of wildfire. Some causes of wildfires are of human origin, whether due to negligence or intentionally, and they take place most of the time along the roads, so the proximity to the road network can increase the risk of ignition.
The road network and forest paths play a fundamental role in terms of both preventing and fighting wildfires, as they allow access to fire-fighting vehicles and can function as a containment strip.
The areas that present the greatest distance to the road network and forest paths are the municipalities of Valpaços, Vila Pouca de Aguiar, Alijó, Vila Real and the northwest part of the Municipality of Montalegre. There is evidence between this greater distance and the relief conditions (higher altitudes and greater slopes) in the municipalities of Montalegre and Vila Pouca de Aguiar. Except for the Municipality of Vila Real, this greater distance to the road network and forest paths is also associated with municipalities with less population density. The predominant class in the district is the 50 to 100 m Euclidean class with 20.74% of the district area. It is important to mention here that most of the district (59.24%, 255 171 ha) has a Euclidean distance to the road network and forest paths up to 150 m. It should also be noticed that 16.00% of the district area (68 908 ha) is more than 500 m from a road.
Regarding the density of the road network and forest paths, expressed in meters of track per hectare (m/ha), being in the municipalities of Alijó, Valpaços, Vila Pouca de Aguiar, Vila Real and northwest of the Municipality of Montalegre, which has a lower density. The density class of the road network and forest paths that predominates in the district is ≤10 m/ha, corresponding to 24.62% of the district area. An area of the district where the density of roads is greater than 70 m/ha corresponds to 14.57% of the total area.
The exposure of the slopes has its influence on the spread of wildfires, by determining the variations in atmospheric weather during the day. As the sun changes its position, so do the temperature and relative humidity of the air, the moisture content of the fuels and the direction and speed of local winds. On the one hand, the slopes exposed to the south and southwest have meteorological conditions and fuel favourable to the rapid inflammation and spread of fire. On the other hand, the slopes exposed to the north and northeast have a higher moisture content of the fuel, burn more slowly, and reach lower temperatures (Almeida et al., 1995).
The predominant exposure orientation in the District of Vila Real is southeast with about 15.19% (65 434 ha) of the district’s area. However, the different exhibits have very similar values, such as the exposure that corresponds to 14.87% of the district area, the southern exposure that assumes 13.20% of the total area of the district and the western exposure that occupies 12.78% of the study territory. The slopes facing the southwest have an occupation area corresponding to 12.01% of the territory, the exposure to the northwest corresponds to 11.35% of the district area and the exposures to the northeast and the north assume, respectively, 10.62 and 9.28% of the total district area. The flat areas are the least significant, assuming 0.7% (3 019 ha) of the district area.
The elaboration of the wildfire risk map (Figure 3) for the District of Vila Real followed the methodology presented in Section 3. The District of Vila Real presents a certain homogeneity in the distribution of risk across the five classes: very low, low, medium, high, and very high. However, the medium risk class is the predominant one, representing 22.13% of the total area of the district, somewhat distributed throughout the territory. The very low risk represents 18.95% of the district area (77 846 ha), with the Municipality of Montalegre standing out in this risk class. The low risk (19.27%) has a significant expression in the Municipality of Chaves, north of the Municipality of Vila Pouca de Aguiar, west of the Municipality of Vila Real, in some areas of the Municipality of Alijó and the centre of the Municipality of Sabrosa. The high risk of wildfire, which represents 80153 ha (19.51%) of the total district, stands out in the centre of Chaves and the southeast in the Municipality of Valpaços. The very high-risk class, representing 20.13% of the district’s area (82 694 ha), is found in the central-west area of the district (Boticas, Ribeira de Pena and Vila Pouca de Aguiar municipalities) and south of the district (Mesão Frio, Santa Marta de Penaguião, Peso da Régua and Sabrosa municipalities). It is possible to verify that all over the municipalities, there is a very high risk of wildfire. It can be said that 39.64% of the district’s area presents a high or very high risk and that most of the district (61.78%, 253 765 ha) has, at least, an average wildfire risk.
Figure 3 Wildfire risk of the District of Vila Real. A full-colour version of this figure can be found on the ICE Virtual Library
The District of Vila Real has at its disposal 26 fire brigades distributed in the 14 municipalities, with the Municipality of Alijó having five fire departments, the Municipality of Chaves having three fire departments, the municipalities of Montalegre, Ribeira de Pena, Sabrosa, Santa Marta de Penaguião, Valpaços and Vila Real having two fire brigades each and the remaining municipalities having one fire brigade each.
Quick intervention is an essential contribution to effectiveness in combating wildfires. The normal fire behaviour has a low rate of spread and is not dependent on the time factor. This type of behaviour does not put firefighters at great risk, contrary to what happens with the extreme behaviour of the fire in which there is a dynamic change in the characteristics of the fire over time, as shown by studies on slopes by Dold and Zinoviev (2009). According to Viegas et al. (2011), extreme fire behaviour is defined as ‘the set of propagation characteristics and properties that call into question the possibility of security control using the available technology and knowledge’. In cases of extreme fire behaviour, the rate of spread of the fire front has a dynamic behaviour and its properties depend not only on the fuel, the weather conditions and the topography, but also on the development of the fire over time, suffering this one acceleration by varying the rate of spread and the energy released (Viegas, 2004; Viegas, 2006; Viegas and Pita, 2004; Xie et al., 2017, 2020). Bearing in mind the importance of the time factor in the development of wildfires, it is essential that the first intervention is as quick as possible, so it is crucial to have a map that provides the travel time from the fire departments.
To provide a decision support tool regarding planning to combat wildfires, the map shown in Figure 4 was drawn up, which refers to the time taken by the fire brigades to reach the location of a wildfire. The travel times were calculated for the road network and previously presented forest paths, assuming an average speed of 50 km/h in the entire network, with a bidirectional route and for a maximum travel time of 1 h and 30 min.
Figure 4 Route times from fire departments of the District of Vila Real. A full-colour version of this figure can be found on the ICE Virtual Library
In the analysis of this parameter, five classes were considered: 0 to 5 min, 5 to 10 min, 10 to 15 min, 15 to 20 min and more than 20 min, with more than 20 min in many cases of wildfires, being too long for the arrival of the first terrestrial means. A wildfire has a complex temporal and spatial evolution, and time is fundamental to the evolution of fire behaviour. The behaviour of fire is dynamic, in the sense that the characteristics of propagation change over time, even with permanent boundary conditions. The interaction of a fire with the environment, through the convection currents generated by the fire itself that modify the surrounding environment, can override the other factors that condition propagation, giving rise to the explicit dependence of time, present since the beginning of the propagation.
Thus, it is essential to reduce the travel time of the fire brigades, predicting and anticipating days that are favourable to the occurrence of wildfires and performing the pre-positioning of firefighters in strategic locations to reduce this travel time.
Analysing Figure 4, it is possible to state that the predominant time class is greater than 20 min, which encompasses 33.26% (142 450 ha) of the District of Vila Real area. Of the total district area, 22.47% is reached in 10 to 15 min, 19.77% in 15 to 20 min and 17.34% in 5 to 10 min, with only 7.16% (30 645 ha) having travel time less than 5 min. It should be noted here that most of the district area (53.03%) has a travel time of more than 15 min.
To provide a decision support tool, this time within the scope of the prevention of wildfires, through the optimisation of surveillance, the map shown in Figure 5 was prepared. This map represents the surveillance priority in the District of Vila Real. As mentioned in Section 2.4, a large part of the district is not visible from any lookout post (31.70% of the total area of the district) or only from one (38.34% of the district area). Thus, the visibility map was crossed with the wildfire risk map, giving rise to the proposal for a surveillance priority map for the District of Vila Real.
Figure 5 Surveillance priorities of the District of Vila Real. A full-colour version of this figure can be found on the ICE Virtual Library.
The analysis carried out here makes it possible to identify areas that are completely hidden (that are not visible from any lookout post) or semi-hidden (that are visible only from one lookout post) and that overlap with areas of high risk or very high risk, indicating the need to carry out greater surveillance by other ways than the network of watch posts available in the District of Vila Real. Regarding other means of surveillance, one can speak, for example, of surveillance by land, or surveillance by air using a remotely piloted aircraft.
In Figure 5, it is possible to see the areas of the district with high or very high risk and that is not visible by any lookout post or that are visible only by a lookout post. Based on this figure, it is possible to claim that throughout the district, there is inadequate visibility from the network of available lookouts and that, taking into account the risk of wildfire, it is necessary to pay special attention to this problem. Once again, the pre-positioning of fire brigades can be an added value, if the combination of pre-positioning of fire brigades to combat and the surveillance mission is done, thus filling this lack of adequate visibility within the study territory.
The District of Vila Real presents, based on the analysis of the wildfire risk, a predominance of average risk in 22.13% of its territory. It should be noted here that 39.64% of the district area has a high or very high risk, so it is necessary to take measures to minimise that risk. All municipalities in the district have areas with a slope greater than 40%, which is a factor that enhances the dynamics of a wildfire.
It is important to mention that most of the territory under study (53.03%) is more than 15 min away from the fire brigades. Thus, it is concluded that it is important to pre-position the fire brigades for earlier first intervention in the areas at greatest risk of fire, to make the combat more effective.
Bearing in mind that a large part of the district is not visible from any lookout post (31.70% of the total area of the district) or only from one (38.34% of the area) that alone does not allow crossing at a point on the map for an exact location, it is essential to reinforce the existing surveillance of the network of watch posts. Therefore, based on the map of surveillance priorities, it is necessary to use other means of surveillance (e.g. by land or by air using a remotely piloted aircraft). The pre-positioning of means that reduces the time of arrival at the event, can also serve to reinforce surveillance and address the poor visibility of the network of lookout posts, especially in areas with a higher risk of wildfire.
It is common knowledge that Portugal has a structural problem of spatial planning, which has been a major threat to the safety of people and their property when major wildfires occurred. In the context of climate change, Portugal could be the country of the European Community that suffers the most from climate change, making the frequency of this type of natural catastrophe more recurrent and with increasing intensity. Data from recent years have shown this reality.
The events of June and October 2017 led to systemic reform in preventing and fighting wildfires. Due to their complexity, their study and management require the involvement of multiple institutions and entities, in a multidisciplinary approach to the problem. Citizen involvement is essential, as the collective awareness that protection and security are the responsibility of all and for all is very important. Making communities more resilient is important for citizens to be able to protect themselves in exceptional situations such as those experienced in 2017.
The execution of this study is particularly important in the operational component, both in terms of prevention and in combating wildfires. All the information presented here is very useful as a decision support system. The execution of this study at the district level makes it possible to standardise methodologies and analyses that at the municipal level it would be different, making here a characterisation of the whole and not just of the parts. All scientific knowledge that is presented here must be disseminated by the operational entities to contribute to better management of the wildfire risk.
As future work, it would be interesting to study the wildfire risk for the District of Vila Real district based on other methods that also involve GIS tools.
Acknowledgements
The first author would like to thank the Foundation for Science and Technology Ministry of Science, Technology and Higher Education (FCT/MCTES/FSE/UE) for the PhD Grant SFRH/BD/138235/2018.
The third and fifth authors would like to thank the Foundation for Science and Technology Ministry of Science, Technology and Higher Education (FCT/MCTES) for the financial support to CESAM (UIDP/50017/2020+UIDB/50017/2020), through national funds.
