European Forest Fire Information System Research Articles

FIREMAP: Cloud-based software to automate the estimation of wildfire-induced ecological impacts and recovery processes using remote sensing techniques

The formulation and planning of integrated fire management strategies must be strengthened by decision support systems about fire-induced ecological impacts and ecosystem recovery processes, particularly in the context of extreme wildfire events that challenge land management initiatives. Wildfire data collection and analysis through remote sensing earth observations is of utmost importance for this purpose. However, the needs of land managers are not always met because the exploitation of the full potential of remote sensing techniques requires a high level of technical expertise. In addition, data acquisition and storage, database management, networking, and computing requirements may present technical difficulties. Here, we present FIREMAP software, which leverages the potential of Google Earth Engine (GEE) cloud-based platform, an intuitive graphical user interface (GUI), and the European Forest Fire Information System (EFFIS) wildfire database for wildfire analyses through remote sensing techniques and data collections. FIREMAP software allows automatic computing of (i) machine learning-based burned area (BA) detection algorithms to facilitate the mapping of (historical) fire perimeters, (ii) fire severity spectral indices, and (iii) post-fire recovery trajectories through the inversion of physically-based radiative transfer models. We introduce (i) the FIREMAP platform architecture and the GUI, (ii) the implementation of well-established algorithms for wildfire science and management in GEE, (iii) the validation of the algorithm implementation in fifteen case-study wildfires across the western Mediterranean Basin, and (iv) the near-future and long-term planned expansion of FIREMAP features.

In temperate Europe, fire is already here: The case of The Netherlands

Landscape fires are usually not associated with temperate Europe, yet not all temperate countries record statistics indicating that actual risks remain unknown. Here we introduce new wildfire statistics for The Netherlands, and summarize significant events and fatalities. The period 2017–2022 saw 611 wildfires and 405 ha burned per year, which Copernicus’ European Forest Fire Information System satellite data vastly underestimate. Fires burned more heathland than forest, were small (mean fire size 1.5 ha), were caused by people, and often burned simultaneously, in Spring and in Summer drought. Suppression, restoration and traffic delays cost 3 M€ year−1. Dozens of significant events illustrate fire has never been away and has major societal impact amidst grave concerns for firefighter safety. Since 1833, 31 fatalities were reported. A legal framework is needed to ensure continuity of recordkeeping, as the core foundation of integrated fire management, to create a baseline for climate change, and to fulfill international reporting requirements.

Assessing different measures of fire risk for Cultural World Heritage Sites

This study aims to assess whether two publicly available sources of fire threats to 346 Cultural World Heritage Sites across Europe substitute or complement each other. By doing so, a novel measure based on information from the UNESCO periodic report II is created and benchmarked against the European forest fire information system (EFFIS) index. The UNESCO periodic report shows that forest fires are perceived as an actual or foreseeable hazard by 40% of the management of Cultural World Heritage Sites in Europe. When the EFFIS index is linked to the UNESCO World Heritage database, it occurs that 48% of these sites are at high risk of fire, 31% at medium risk and 21% at low risk. Results based on Probit and Fractional Probit estimations reveal that the perceived fire risk relates to several site characteristics as well as location. The regressions using the EFFIS index as dependent variable show indifference to site characteristics even if location is of importance. Estimations give that the perceived fire risk is highest for sites in the East and the North of Europe, while the results for the EFFIS index lead to a dominant risk in the South. A 10° increase in latitude (corresponding to the distances between Vienna and Stockholm or Athens and Vienna) leads to a considerable decrease in the proportion of high fire risk by 28 percentage points (with a sample mean of 48%). Thus, the two measures of fire risks complement rather than substitute each other. Latitude is of no importance for the site managers, although the EFFIS gives this aspect a heavy weight, with low or zero risks in locations at higher latitudes (Iceland, Ireland, Latvia, and Norway) and larger risks in Southern Europe (Cyprus, Malta, Portugal, Albania, Spain, and Greece). In addition, the perception of (wild)-fire threat is significantly lower for cities.

Near real-time generation of a country-level burned area database for Italy from Sentinel-2 data and active fire detections

The first national database of burned areas in Italy at 20 m spatial resolution is presented. It was generated in near real-time by processing, as soon as they were available, all the Sentinel-2 images covering the Italian territory throughout the fire seasons of years 2019–2021. An updated version of the AUTOmatic Burned Area Mapper (AUTOBAM), originally developed in a previous study, was used to produce the database. Active fire data and a feature extraction procedure were used by AUTOBAM to associate a set of quality flags to each burned area.The AUTOBAM product was evaluated by comparing it with reference polygons derived from ground surveys. Even the database derived from the European Forest Fire Information System (EFFIS) was compared to reference polygons to assess the utility of the AUTOBAM product with respect to an already existing one. Good results were obtained by AUTOBAM in terms of detection capability (the Correctness parameter) and overlap factor (both larger than 60%). However, quite high values of the commission error were obtained, especially in 2019. Through a per land cover analysis, it was found that this error mostly occurred in cultivated land. Excluding the latter target, the commission error was always less than 35%, the omission error was less than 27% and the Dice Coefficient was larger than 69%.

Cross-country risk quantification of extreme wildfires in Mediterranean Europe.

We estimate the country-level risk of extreme wildfires defined by burned area (BA) for Mediterranean Europe and carry out a cross-country comparison. To this end, we avail of the European Forest Fire Information System (EFFIS) geospatial data from 2006 to 2019 to perform an extreme value analysis. More specifically, we apply a point process characterization of wildfire extremes using maximum likelihood estimation. By modeling covariates, we also evaluate potential trends and correlations with commonly known factors that drive or affect wildfire occurrence, such as the Fire Weather Index as a proxy for meteorological conditions, population density, land cover type, and seasonality. We find that the highest risk of extreme wildfires is in Portugal (PT), followed by Greece (GR), Spain (ES), and Italy (IT) with a 10-year BA return level of 50'338 ha, 33'242 ha, 25'165 ha, and 8'966 ha, respectively. Coupling our results with existing estimates of the monetary impact of large wildfires suggests expected losses of 162-439 million € (PT), 81-219 million € (ES), 41-290 million € (GR), and 18-78 million € (IT) for such 10-year return period events. SUMMARY: We model the risk of extreme wildfires for Italy, Greece, Portugal, and Spain in form of burned area return levels, compare them, and estimate expectedlosses.

Anthropogenic and Lightning Fire Incidence and Burned Area in Europe

Fires can have an anthropogenic or natural origin. The most frequent natural fire cause is lightning. Since anthropogenic and lightning fires have different climatic and socio-economic drivers, it is important to distinguish between these different fire causes. We developed random forest models that predict the fraction of anthropogenic and lightning fire incidences, and their burned area, at the level of the Nomenclature des Unités Territoriales Statistiques level 3 (NUTS3) for Europe. The models were calibrated using the centered log-ratio of fire incidence and burned area reference data from the European Forest Fire Information System. After a correlation analysis, the population density, fractional human land impact, elevation and burned area coefficient of variation—a measure of interannual variability in burned area—were selected as predictor variables in the models. After parameter tuning and running the models with several train-validate compositions, we found that the vast majority of fires and burned area in Europe has an anthropogenic cause, while lightning plays a significant role in the remote northern regions of Scandinavia. Combining our results with burned area data from the Moderate Resolution Imaging Spectroradiometer, we estimated that 96.5 ± 0.9% of the burned area in Europe has an anthropogenic cause. Our spatially explicit fire cause attribution model demonstrates the spatial variability between anthropogenic and lightning fires and their burned area over Europe and could be used to improve predictive fire models by accounting for fire cause.

Automatic forest fire danger rating calibration: Exploring clustering techniques for regionally customizable fire danger classification

Effective wildfire management begins with fire prevention and the assessment of the local fire danger. Typical approaches to fire danger classification rely heavily on manual analysis for specific regions based on expert knowledge. In this paper, a novel approach is proposed for the automatic calibration of fire danger classes based on the Canadian Forest Fire Weather Index System (CFFWIS) applied for specific regions. The proposed automatic calibration method is based on clustering algorithms, namely k-means, fuzzy c-means, Gaussian mixture models, and data-clouds, which are used to identify clusters in datasets composed of elements from CFFWIS and wildfire historical records. The clusters are associated with fire danger classes which are separated by proposed thresholds based on the fire weather index values contained within each cluster. Exhaustive experiments ensured an accurate comparison of performance with the analysis of our fire danger classes against the classes defined by the European Forest Fire Information System (EFFIS), also based on the CFFWIS. These experiments consider individual information from each of the selected European regions from a total of 769 regions with available data, with validation aimed at the analysis of the large fires in a general context, and a case study for Portuguese regions.

Impact of large wildfires on PM<sub>10</sub> levels and human mortality in Portugal

Abstract. Uncontrolled wildfires have a substantial impact on the environment, the economy and local populations. According to the European Forest Fire Information System (EFFIS), between 2000 and 2013 wildfires burned up to 740 000 ha of land annually in the south of Europe, Portugal being the country with the highest percentage of burned area per square kilometre. However, there is still a lack of knowledge regarding the impacts of the wildfire-related pollutants on the mortality of the country's population. All wildfires occurring during the fire season (June–July–August–September) from 2001 and 2016 were identified, and those with a burned area above 1000 ha (large fires) were considered for the study. During the studied period (2001–2016), more than 2 million ha of forest (929 766 ha from June to September alone) were burned in mainland Portugal. Although large fires only represent less than 1 % of the number of total fires, in terms of burned area their contribution is 46 % (53 % from June to September). To assess the spatial impact of the wildfires, burned areas in each region of Portugal were correlated with PM10 concentrations measured at nearby background air quality monitoring stations. Associations between PM10 and all-cause (excluding injuries, poisoning and external causes) and cause-specific mortality (circulatory and respiratory) were studied for the affected populations using Poisson regression models. A significant positive correlation between burned area and PM10 was found in some regions of Portugal, as well as a significant association between PM10 concentrations and mortality, these being apparently related to large wildfires in some of the regions. The north, centre and inland of Portugal are the most affected areas. The high temperatures and long episodes of drought expected in the future will increase the probabilities of extreme events and therefore the occurrence of wildfires.

EFFECTS OF GEOSPATIAL DATA SOURCES ON THE IDENTIFICATION AND CHARACTERIZATION OF BURNT AREAS IN PORTUGAL

Abstract. The work presented in this paper compares the burnt areas in continental Portugal in 2017 and 2018 mapped by three initiatives, namely the Portuguese Institute of Nature and Forests Conservation (ICNF), the Corine Land Cover (CLC) inventory of the Copernicus programme and the European Forest Fire Information System (EFFIS). Then, the Land Use Land Cover (LULC) classes affected by the 2017 burnt areas mapped by ICNF are analysed considering CLC 2018 and the 2018 LULC map produced by the Portuguese National Mapping Agency (Direção Geral do Território) – “Carta de Ocupação do Solo” (COS 2018). To enable a comparison between the classes of both LULC products, a nomenclature was selected and both CLC 2018 and COS 2018 were mapped into the chosen classes. The comparison of the burnt area’s extent showed that there are large differences in both area and levels of detail between the analysed data sources. The results regarding the LULC classes affected by the 2017 fires mapped by ICNF show large differences in terms of burnt area in each class as well as the proportion of the burnt areas associated to the classes. This analysis shows that very different results may be reached if different products are used, and therefore a large level of uncertainty is associated with the conclusions achieved with these products.

Current Wildland Fire Patterns and Challenges in Europe: A Synthesis of National Perspectives

Changes in climate, land use, and land management impact the occurrence and severity of wildland fires in many parts of the world. This is particularly evident in Europe, where ongoing changes in land use have strongly modified fire patterns over the last decades. Although satellite data by the European Forest Fire Information System provide large-scale wildland fire statistics across European countries, there is still a crucial need to collect and summarize in-depth local analysis and understanding of the wildland fire condition and associated challenges across Europe. This article aims to provide a general overview of the current wildland fire patterns and challenges as perceived by national representatives, supplemented by national fire statistics (2009–2018) across Europe. For each of the 31 countries included, we present a perspective authored by scientists or practitioners from each respective country, representing a wide range of disciplines and cultural backgrounds. The authors were selected from members of the COST Action “Fire and the Earth System: Science & Society” funded by the European Commission with the aim to share knowledge and improve communication about wildland fire. Where relevant, a brief overview of key studies, particular wildland fire challenges a country is facing, and an overview of notable recent fire events are also presented. Key perceived challenges included (1) the lack of consistent and detailed records for wildland fire events, within and across countries, (2) an increase in wildland fires that pose a risk to properties and human life due to high population densities and sprawl into forested regions, and (3) the view that, irrespective of changes in management, climate change is likely to increase the frequency and impact of wildland fires in the coming decades. Addressing challenge (1) will not only be valuable in advancing national and pan-European wildland fire management strategies, but also in evaluating perceptions (2) and (3) against more robust quantitative evidence.

A methodology to estimate forest fires burned areas and burn severity degrees using Sentinel-2 data. Application to the October 2017 fires in the Iberian Peninsula

A methodology to estimate the extent of areas affected by forest fires, as well as the burn severity levels using Sentinel 2 images (10 and 20 m) is proposed and applied to the fires occurred in October 2017 in Spain and Portugal. An extension larger than 250,000 ha and 4 burn severity levels (low, moderate, high and very high) have been obtained. The comparison with the European Forest Fire Information System (EFFIS), which uses MODIS images (250 m), shows that the methodology improves the area estimate by 10 % in commission area. In terms of burn severity levels, the Separability index (SI) and the Kappa statistic (k) show a high correlation between Sentinel-2 and EFFIS (SI values higher than one in all cases and k higher than 0.69, respectively).

ERA5-based global meteorological wildfire danger maps

Forest fires are an integral part of the natural Earth system dynamics, however they are becoming more devastating and less predictable as anthropogenic climate change exacerbates their impacts. In order to advance fire science, fire danger reanalysis products can be used as proxy for fire weather observations with the advantage of being homogeneously distributed both in space and time. This manuscript describes a reanalysis dataset of fire danger indices based on the Canadian Fire Weather Index system and the ECMWF ERA5 reanalysis dataset, which supersedes the previous dataset based on ERA-Interim. The new fire danger reanalysis dataset provides a number of benefits compared to the one based on ERA-Interim: it relies on better estimates of precipitation, evaporation and soil moisture, it is available in a deterministic form as well as a probabilistic ensemble and it is characterised by a considerably higher spatial resolution. It is a valuable resource for forestry agencies and scientists in the field of wildfire danger modeling and beyond. The global dataset is produced by ECMWF, as the computational centre of the European Forest Fire information System (EFFIS) of the Copernicus Emergency Management Service, and it is made available free of charge through the Climate Data Store.

Weather-related thresholds for wildfire danger in a Mediterranean region: The case of Greece

The objective of the following paper is to identify weather-related thresholds for wildfire danger, i.e., the potential extent of fire, based on local weather conditions. The target area is Greece, a wildfire prone Mediterranean country which experienced on average 2000 wildfires annually over the last two decades. Initially, the Fire Weather Index (FWI) component of the Canadian Fire Weather Index (FWI) System (FWI System) adopted by the European Forest Fire Information System (EFFIS), is evaluated with respect to its wildfire danger predictive ability. Hence, weather and wildfire data at municipality level and on a daily basis, for the period 2000–2016 are exploited. The analysis showed that the FWI thresholds proposed by EFFIS for assessing the level of fire danger in Europe are too low for the case of Greece and, therefore, are not representative of the country's fire weather conditions. Two statistical approaches, cluster analysis and non-linear least-squares regression, are subsequently applied to determine the most appropriate FWI thresholds for discrete levels of wildfire danger. The results are presented in 4 sets of FWI thresholds and they are further evaluated on the basis of verification measures. All sets of FWI thresholds were found to significantly improve the predictability of wildfire danger compared to the EFFIS fire danger class thresholds. In particular, two of them were found to meet selected performance requirements for a balanced predictive performance, namely a reduced overestimation of wildfire danger and increased reliability in danger classification. The results are expected to have significant practical implications for wildfire prevention and risk mitigation strategies implemented by the forest fire control agencies of the country.

Wildfire burnt area patterns and trends in Western Mediterranean Europe via the application of a concentration index

AbstractThe most widely used metrics to characterize wildfire regime and estimate the impact of wildfires are total burnt area (BA) and the number of fire events (FE). However, these are insufficient to analyse the threat to society of a new fire regime characterized by a higher occurrence of very large events. To overcome this weakness, we propose the use of a Concentration Index (CIB) which makes it possible to identify spatio‐temporal patterns. The frequency distribution of BA follows a negative exponential distribution almost everywhere, in which a small minority of FE is responsible of the majority of BA. In this article, the spatio‐temporal behaviour of BA is analysed in Western Mediterranean Europe, with particular focus on Portugal, Spain, France and Italy, using data from the European Forest Fire Information System and national wildfire databases. This is the first time that the CI has been applied to wildfire events. This research shows that, in most Mediterranean European countries, the amount of BA is increasingly related with a lower number of fires. The spatio‐temporal distribution of CIB shows high variability in all of the countries analysed in Europe. Portugal and Spain show increasing significant trends of CIB + 7.6% (p‐value = 0.001) and + 1.3% per decade (p‐value = 0.003). Statistically significant correlations for Portugal, Spain and Italy are also found between the annual CIB and several teleconnection indices. The application of the CIB demonstrates its discriminatory ability, which is a key point in detecting vulnerable areas and temporal trends under climate change.

Comparative analysis of CORINE and climate change initiative land cover maps in Europe: Implications for wildfire occurrence estimation at regional and local scales

Updated and harmonized land cover (LC) data is essential for wildfire estimation in fire-prone areas as is the case in southern Europe. CORINE Land cover (CLC) and ESA Climate Change Initiative Land Cover (CCI-LC) maps have been analyzed and compared their performance in the estimation of wildfire occurrence in Europe at regional and local scales for the period 2010–2014. LC maps legends were harmonized and similarities and discrepancies were compared. Overall agreement between the two maps for the whole Europe was ˜75%. Forest and agriculture showed the largest agreement, while shrubland and grassland the lowest. Quantity and allocation disagreements were calculated including exchange and shift components (Pontius and Santacruz, 2014) which provided detailed information about the contribution of each class to the overall disagreement. Spatial discrepancies were found in areas where grassland and shrubland were the dominant classes as in United Kingdom or East Turkey. Land Use and Coverage Area frame Survey (LUCAS) was used as ground truth for validation purposes. The agreement with LUCAS was slightly higher for CCI-LC (59%) than for CLC (56%). Generalized Linear Models (GLM), based on presence-absence of wildfires, were used to estimate wildfire occurrence at 3 × 3 km grid cell resolution from both LC maps at the European scale. LC interfaces and climatic variables (temperature and precipitation) where used as explicative variables while fires from European Forest Fire Information System EFFIS (2010–2014 period) were used as response variable. Wildfire occurrence was also estimated with the two maps at local scale in a test region (Zamora, Spain) using a more precise location of the response variable (x, y fire ignition points). At the European scale models obtained within the two maps showed similar results. CCI-LC model sensitivity was 77.26%, specificity 25.89% and omission error 22.74% while CLC model sensitivity was 75.68%, specificity 29.99% and omission error 24.32%. However, CLC performed slightly better in terms of the percent correct classification (62%). In the test region the models achieved better results in terms of specificity (66.07% and 68.93% for CCI-LC and CLC models respectively) and percent correct classification (˜68% for CLC model). At local scale CLC model performed better than CCI-LC model. Wildfire occurrence estimation was more accurate at local scale because of the differences in the spatial accuracy of the response variable used.

Appraisal of the Sentinel-1 & 2 use in a large-scale wildfire assessment: A case study from Portugal's fires of 2017

The recent launch of Sentinel missions offers a unique opportunity to assess the impacts of wildfires at higher spatial and spectral resolution provided by those Earth Observing (EO) systems. Herein, an assessment of the Sentinel-1 & 2 to map burnt areas has been conducted. Initially the use of Sentinel-2 solely was explored, and then in combination with Sentinel-1 and derived radiometric indices. As a case study, the large wildfire occurred in Pedrógão Grande, Portugal in 2017 was used. Burnt area estimates from the European Forest Fires Information System (EFFIS) were used as reference. Burnt area was delineated using the Maximum Likelihood (ML) and Support Vector Machines (SVMs) classifiers, and two multi-index methods. Following this, burn severity was assessed using SVMs and Artificial Neural Networks (ANNs), again for both standalone Sentinel-2 data and in combination with Sentinel-1 and spectral indices. Soil erosion predictions were evaluated using the Revised Universal Soil Loss Equation (RUSLE) model. The effect of the land cover derived from CORINE operational product was also evaluated across the burnt area and severity maps. SVMs produced the most accurate burnt area map, resulting a 94.8% overall accuracy and a Kappa coefficient of 0.90. SVMs also achieved the highest accuracy in burn severity levels estimation, with an overall accuracy of 77.9% and a kappa of 0.710. From an algorithmic perspective, implementation of the techniques applied herein, is based on EO imagery analysis provided nowadays globally at no cost. It is also robust and adaptable, being potentially integrated with other high EO data available. All in all, our study contributes to the understanding of Mediterranean landscape dynamics and corroborates the usefulness of Sentinels data in wildfire studies.

Fire danger rating over Mediterranean Europe based on fire radiative power derived from Meteosat

Abstract. We present a procedure that allows the operational generation of daily forecasts of fire danger over Mediterranean Europe. The procedure combines historical information about radiative energy released by fire events with daily meteorological forecasts, as provided by the Satellite Application Facility for Land Surface Analysis (LSA SAF) and the European Centre for Medium-Range Weather Forecasts (ECMWF). Fire danger is estimated based on daily probabilities of exceedance of daily energy released by fires occurring at the pixel level. Daily probability considers meteorological factors by means of the Canadian Fire Weather Index (FWI) and is estimated using a daily model based on a generalized Pareto distribution. Five classes of fire danger are then associated with daily probability estimated by the daily model. The model is calibrated using 13 years of data (2004–2016) and validated against the period of January–September 2017. Results obtained show that about 72 % of events releasing daily energy above 10 000 GJ belong to the “extreme” class of fire danger, a considerably high fraction that is more than 1.5 times the values obtained when using the currently operational Fire Danger Forecast module of the European Forest Fire Information System (EFFIS) or the Fire Risk Map (FRM) product disseminated by the LSA SAF. Besides assisting in wildfire management, the procedure is expected to help in decision making on prescribed burning within the framework of agricultural and forest management practices.

Caliver: An R package for CALIbration and VERification of forest fire gridded model outputs.

The name caliver stands for CALIbration and VERification of forest fire gridded model outputs. This is a package developed for the R programming language and available under an APACHE-2 license from a public repository. In this paper we describe the functionalities of the package and give examples using publicly available datasets. Fire danger model outputs are taken from the modeling components of the European Forest Fire Information System (EFFIS) and observed burned areas from the Global Fire Emission Database (GFED). Complete documentation, including a vignette, is also available within the package.

A Fully Automatic Instantaneous Fire Hotspot Detection Processor Based on AVHRR Imagery—A TIMELINE Thematic Processor

The German Aerospace Center’s (DLR) TIMELINE project aims to develop an operational processing and data management environment to process 30 years of National Oceanic and Atmospheric Administration (NOAA)—Advanced Very High Resolution Radiometer (AVHRR) raw data into L1b, L2 and L3 products. This article presents the current status of the fully automated L2 active fire hotspot detection processor, which is based on single-temporal datasets in orbit geometry. Three different probability levels of fire detection are provided. The results of the hotspot processor were tested with simulated fire data. Moreover, the processing results of real AVHRR imagery were validated with five different datasets: MODIS hotspots, visually confirmed MODIS hotspots, fire-news data from the European Forest Fire Information System (EFFIS), burnt area mapping of the Copernicus Emergency Management Service (EMS) and data of the Piedmont fire database.

Regional aspects of modelling burned areas in Europe

This paper presents a series of improvements to the quantitative modelling of burned areas in Europe under historical climate. The Standalone Fire Model (SFM) based on a state-of-the-art large scale mechanistic fire modelling algorithm is used to reproduce historical burned areas reported in the two publicly available datasets – European Forest Fire Information System (EFFIS) and Global Fire Emissions Database (GFED). The most recent versions of these sources allow a broader validation of SFM’s modelled burned areas at a country level. Our analysis is carried out for the years 2000–2008 for 17 European countries utilising both EFFIS and GFED datasets for model benchmarking. We suggest improving the original model by modifying the fire probability function reflecting fuel moisture. This modification allows for a dramatic improvement of accuracy in modelled burned areas for a range of European countries. We also explore in detail a pixel-level parametrisation of firefighting efficiency in SFM along with modifications of the biomass map. In comparison with the aggregated country-level approach, the advantages of the finer calibration are quite minor for the most recent version of the GFED dataset. Overall, the annual burned areas modelled by this improved SFM version are in good agreement with historical observations.