Fire Module
This document provides a thematical background for the topic of wildfires, including examples from Africa and the context of Earth Observation.

Introduction
Fires (or Wildfires) are considered one of the most destructive natural disasters that can cause adverse impacts to nature and society, as well as economy and built environments. Generally, fires are ignited by natural or human activities. Natural activities include lightning and volcanic eruptions, whereas human activities basically include arson, campfires, and unextinguished cigarettes.

Approximately 10% of all fires in the global landscape are ignited by natural causes, mainly lightning. The remaining 90% of fires are caused by humans [2]. According to a report in the U.S., data in the 2000 – 2017 period revealed that nearly 85% of the wildfires across the U.S. are caused by human activities [3], [4]. In Africa, EM-DAT data revealed a total of 471 deaths and 181 417 affected people in 36 wildfires from 1973 to 2022 [5].

Fires can cause fatalities, and serious injuries – high damage to residential areas and workplaces, as well as public infrastructures. Further, increased wildfire activity can act as a driver of climate change via the increased release of CO², soot, and other aerosols during combustion, and through the removal of vegetation which would otherwise have served as a CO² sink [6]. Even though fewer deaths occur in the wildfire hazards than in floods or earthquakes [5], [7], it still stands as a destructive hazard, a threat to nature and society. Wildfires can be classified based on different characteristics, Figure 2 gives an overview of different types of Wildfires.

Every year; fires ravage forests, farmlands, other natural environments, and also built-up areas and cause billions of Euros worth of damage across the globe. According to a report for the year 2014 by the Fire Protection Research Foundation, the total cost of fire - defined as the collective of all net expenditure on fire protection and all net losses due to fire incidents - in the United States is 328.5 billion USD, which was 1.9% of the country’s GDP [9]. Specifically, fighting “wildfires” in the United States costs over 3 billions of dollars annually [10], [11]. On the other hand, Canadian wildland fire management agencies annually invest around 800 million to 1.4 billion CAD to protect citizens, private residences, businesses, wood supply, and critical infrastructure [12]. Meanwhile, in the EU, statistics from 2019 revealed that the 27 member states spent a combined 30.9 billion Euros on fire-protection services [13], [14]. If we ought to consider these costs globally, hundreds of billions of Euros are spent on fire preparedness, mitigation, management, and recovery.

Aside from damage to nature, economy and impact on humans, experiencing a wildfire can also arise several medical and psychiatric disorders. Post-traumatic stress disorder, depression, and anxiety are the most common mental disorders in the post-fire period, along with various medical conditions such as hypertension, gastrointestinal disorders, diabetes, and chronic obstructive pulmonary disease and asthma exacerbation [15].

Wildfires in Africa – An Overview

Apart from the wildfire statistics in Africa, EM-DAT data also revealed that 32 714 people were killed in 1 182 flood events between 1927 and 2022, and 21 585 lost their lives in 74 earthquake events from 1910 to 2020 [16]. This data clearly confirms the thesis that fewer losses occur in wildfire events than in floods or earthquakes, but it doesn’t include the actual damage to nature, climate – structural damage to the built-up areas and financial damage to the governments.

A total of 17 countries in Africa were affected by the wildfires from 1973 to 2022. South Africa is in the first place with 10 wildfire events, followed by Algeria with 4 and Benin, Central African Republic and the Democratic Republic of the Congo with 3 wildfire events each [16]. Statistics by country is presented in Table 1.

The Role of Remote Sensing and Earth Observation

Remote sensing techniques are a highly feasible and effective tool for describing patterns of the occurrence of fire in various ecosystems [17]. These techniques provide a means for analyzing conditions and monitoring changes over large geographic extents [6]. Data acquired by space-based remote sensing instruments (See Figure 3 for an illustration of the performed by the Sentinel-2 satellite) can help authorities, scientists and individuals to monitor, manage and prevent wildfires at local, regional and global scales as freely available Earth Observation (EO) data allows the depicted user groups to overcome various stages of the hazard.

In this regard; monitoring environmental properties (meteorological and vegetation-related) with EO data and detecting areas with a possible wildfire risk can help authorities take measures and better manage the event, before it happens or gets too serious.

Fire danger/hazard assessment is used to reveal potential fires and is accomplished through two primary methods: point-wise meteorological data, and remote sensing technologies & GIS [6]. In the meteorological method, it’s possible to assess fire danger with various operating systems, such as Fire Weather Index (FWI) or McArthur’s Forest Fire Danger Rating System (FFDRS) by using temperature, precipitation, humidity, and wind speed data [6]. Remote sensing based approaches are also commonly used. Fire risk maps produced using remote sensing approaches include data such as land cover, fuel type, topography (elevation, slope, aspect), and various vegetation indices - such as NDMI (Normalized Difference Moisture Index), and NDVI (Normalized difference Vegetation Index) – and in some cases EVI (Enhanced Vegetation Index), and SAVI (Soil-Adjusted Vegetation Index) [6], [16].

It’s also possible to detect active fires with remote sensing. It’s one of the key points as it has a significant effect on the scope of managing and preventing the possible danger, as well as alerting the public to wildfires to minimize associated negative impacts [6]. Moderate Resolution Imaging Spectroradiometer (MODIS), on board the EOS-AM (Terra) and EOS-PM (Aqua) satellites, and Visible Infrared Imaging Radiometer Suite (VIIRS) are among the most used remote sensing platforms for active fire detection [6], [19], [20]. MODIS Terra and Aqua have become the primary sensors for active fire detection through their high temporal resolution and ability to detect fires, using their special channels designed for fire monitoring [6]. On the Global Wildfire Information System (GWIS), European Forest Fire Information System (EFFIS) and Fire Information for Resource Management System (FIRMS), both MODIS and VIIRS platforms are used since VIIRS complements MODIS and provides a better spatial resolution (375m), as compared to MODIS (1km) [19], [21], [22].

EO data is also powerful during the fire event as no physical intervention is needed to monitor the current situation near real time, analyze how the danger might extend, and better plan the extinguishing and rescuing processes. This can be done by aforementioned active fire detection processes, along with burned area estimates and burn severity assessments.

Burn severity assessments during the fire event and in post-fire process help authorities to determine the extent of the fire and its impacts on the vegetation. Generally, NDVI and NBR (Normalized Burn Ratio) are commonly used spectral indices for burn severity assessments. Recently, NBR – which is calculated using NIR (Near-Infrared) and SWIR (Shortwave-Infrared) bands (Equation ( 1 ) ) - has replaced NDVI and become the standard index for this purpose [6]. Additionally, NBR is particularly sensitive to the changes in the amount of live green vegetation, moisture content, and some soil conditions which may occur after fire [23]. In this scope, pre-fire and post-fire NBRs are calculated for dNBR (Difference/Delta NBR) – then dNBR (Equation ( 2 ) ) reveals severity in 5 levels: unburned, low, low-moderate, moderate-high, and high severity.

EO data, as well as advanced image analysis techniques, e.g. NDVI time series, SAVI (Soil-Adjusted Vegetation Index), provide an opportunity to evaluate patterns of forest regeneration and vegetation recovery in post-fire processes [25], [26]. The evaluation of vegetation change over time can also be conducted by performing traditional field observations [17], [27]. However, working with remote sensing data would be ideal in large areas and/or where the topography is kind of a problem – considering the costs and the amount of time it would take for field observations. One of the key points for vegetation recovery is the sensor and its spatial resolution used through this process as it will not be possible to detect vegetation in the early stages of recovery with poor (>100 m) spatial resolution. This also applies to burn severity assessments. In this case, it will not be possible to detect burn severity patterns with poor spatial resolution. For this reason, most studies use sensors such as those onboard the Landsat satellites, the Sentinel-2 satellites, and ASTER [6].

Recent Wildfires in Africa

a)    2017 South Africa Wildfires

In June 2017, a large storm referred to as the “Cape Storm” hit the southern coast of South Africa. The winds caused by the storm fueled around 20 – 30 significant fires in the town of Knysna and neighboring areas. Between June 6 and 10, fires killed 7 people and displaced around 10,000 people, and destroyed more than 600 structures in Knysna and neighboring areas. Moreover, the estimated worth of damage to the public infrastructure was 136 million Rands (*9 million Euros), whereas the worth of damage to private property was estimated as 5 billion Rands (*345 million Euros) [28]. Experts later found out that the fires started as a result of lightning [29].

*the rate of 10 June 2017, as 1R = 0,069€, used for the calculation.

b)   2021 Algeria Wildfires

In August 2021, Algeria faced severe wildfires across the Kabylia region, placed in the northern part of the country. The wildfires affected 21 wilayas (states, or provinces), injured more than 200 people, and killed a total of 90 people – including 57 civilians, and 33 soldiers who died during the rescue operations. According to a report by the Algerian Directorate General of Forests, more than 100 000 hectares of forest and farmland were ravaged in the event [30]. On 11 August 2021, Copernicus Emergency Management Service (EMS) was activated over the wildfires in Algeria, and a total of 16 maps were produced in this regard [31]. A general map for the affected areas that includes burnt areas, active fires, and alerts was produced to display the status as of 11 August (See Figure 4). Additionally, a detailed map of Tizi Ouzou area on 20 August, which shows a damage grade assessment, can be viewed here.

c)    2022 Algeria Wildfires

A hazardous fire event occurred in Algeria once more in August 2022, in the northern part of the country. Between 17 and 21 August, when the fires were the strongest, 43 people were killed, more than 200 people injured and around 2 000 people were displaced – additionally, the fires destroyed 10 000 hectares of forest, as well as a zoo. On 18-19 August, the Global Wildfire Information System (GWIS), reported several active fires across northern Algeria. According to GWIS, the fire danger forecast on 19 - 21 August was expected to be from very high to extreme over most of northern Algeria.

References
[1]    C. Cook, “Foto zum Thema Silhouette von 2 Personen, die bei Sonnenuntergang auf dem Rasen stehen – Kostenloses Bild zu Feuer auf Unsplash.” 2021. Accessed: Mar. 05, 2023. [Online]. Available: https://unsplash.com/de/fotos/syuox8fipX4

[2]     FAO Committee on Forestry, “Forests Fires and the Global Fire Platform, Twenty-sixth Session.” 2022. [Online]. Available: https://www.fao.org/3/nj847en/nj847en.pdf

[3]     K. C. Short, “Spatial wildfire occurrence data for the United States, 1992-2018 [FPA_FOD_20210617] (5th Edition), U.S. Department of Agriculture.” 2021. doi: 10.2737/RDS-2013-0009.5.

[4]     U.S. National Park Service, “Wildfire Causes and Evaluations (U.S. National Park Service),” 2022. https://www.nps.gov/articles/wildfire-causes-and-evaluation.htm (accessed Mar. 05, 2023).

[5]     CRED, “EM-DAT | The international disasters database. Centre for Research on the Epidemiology of Disasters,” 2019. https://www.emdat.be/ (accessed Sep. 29, 2019).

[6]     D. M. Szpakowski and J. L. R. Jensen, “A Review of the Applications of Remote Sensing in Fire Ecology,” Remote Sens., vol. 11, no. 22, Art. no. 22, Jan. 2019, doi: 10.3390/rs11222638.

[7]     S. H. Doerr and C. Santín, “Global trends in wildfire and its impacts: perceptions versus realities in a changing world,” Philos. Trans. R. Soc. B Biol. Sci., vol. 371, no. 1696, p. 20150345, Jun. 2016, doi: 10.1098/rstb.2015.0345.

[8]     NASA ARSET Team, “Satellite Observations and Tools for Fire Risk, Detection, and Analysis.” 2021. [Online]. Available: https://appliedsciences.nasa.gov/sites/default/files/2021-05/Fire_Part1.pdf

[9]     J. Zhuang, V. M. Payyappalli, A. Behrendt, and K. Lukasiewicz, Total cost of fire in the United States. Fire Protection Research Foundation Buffalo, NY, USA, 2017.

[10]   J. K. Balch, B. A. Bradley, J. T. Abatzoglou, R. C. Nagy, E. J. Fusco, and A. L. Mahood, “Human-started wildfires expand the fire niche across the United States,” Proc. Natl. Acad. Sci., vol. 114, no. 11, pp. 2946–2951, Mar. 2017, doi: 10.1073/pnas.1617394114.

[11]   A. Parijanlar and C. Welch, “What are the costs of fighting wildfires?,” Megafires Student Generated Pages, 2013. https://serc.carleton.edu/NZFires/megafires/cost.html (accessed Mar. 05, 2023).

[12]   Government of Canada, “Cost of wildland fire protection,” 2021. https://natural-resources.canada.ca/climate-change/impacts-adaptations/climate-change-impacts-forests/forest-change-indicators/cost-fire-protection/17783 (accessed Mar. 05, 2023).

[13]   Eurostat, “How much do governments spend on fire-protection?,” 2021. https://ec.europa.eu/eurostat/web/products-eurostat-news/-/edn-20210504-1 (accessed Mar. 05, 2023).

[14]   A. Tidey, “EU member states spend 0.5% of their money on fire protection services. Euronews,” euronews, Jul. 18, 2022. https://www.euronews.com/my-europe/2022/07/18/eu-member-states-spend-just-05-of-their-money-on-fire-protection-services (accessed Mar. 05, 2023).

[15]   P. To, E. Eboreime, and V. I. O. Agyapong, “The Impact of Wildfires on Mental Health: A Scoping Review,” Behav. Sci., vol. 11, no. 9, Art. no. 9, Sep. 2021, doi: 10.3390/bs11090126.

[16]   B. Yu et al., “Fire risk prediction using remote sensed products: A case of Cambodia,” Photogramm Eng Remote Sens, vol. 83, no. 1, pp. 19–25, 2017.

[17]   S. M. B. dos Santos, A. Bento-Gonçalves, and A. Vieira, “Research on Wildfires and Remote Sensing in the Last Three Decades: A Bibliometric Analysis,” Forests, vol. 12, no. 5, Art. no. 5, May 2021, doi: 10.3390/f12050604.

[18]   AIRBUS, “Happy birthday Sentinel-2A: Five Years of Environmental Insights | Airbus,” Oct. 28, 2021. https://www.airbus.com/en/newsroom/press-releases/2020-06-happy-birthday-sentinel-2a-five-years-of-environmental-insights (accessed Mar. 05, 2023).

[19]   W. Schroeder, P. Oliva, L. Giglio, and I. A. Csiszar, “The New VIIRS 375m active fire detection data product: Algorithm description and initial assessment,” Remote Sens. Environ., vol. 143, pp. 85–96, Mar. 2014, doi: 10.1016/j.rse.2013.12.008.

[20]   W. Schroeder et al., “Validation of GOES and MODIS active fire detection products using ASTER and ETM+ data,” Remote Sens. Environ., vol. 112, no. 5, pp. 2711–2726, May 2008, doi: 10.1016/j.rse.2008.01.005.

[21]   Joint Research Centre, “GWIS - Active Fire Detection. European Commission.” https://gwis.jrc.ec.europa.eu/about-gwis/technical-background/active-fire-detection (accessed Mar. 05, 2023).

[22]   W. Schroeder and L. Giglio, “Visible infrared imaging radiometer suite (VIIRS) 375 m$\backslash$& 750 m active fire detection data sets based on NASA VIIRS land science investigator processing system ${$(SIPS)$}$ reprocessed data-Version 1 product User’s guide Version 1.2 available at: https://lpdaac. usgs. gov/documents/132.” VNP14_User_Guide_v1, 2017.

[23]   J. D. Miller and A. E. Thode, “Quantifying burn severity in a heterogeneous landscape with a relative version of the delta Normalized Burn Ratio (dNBR),” Remote Sens. Environ., vol. 109, no. 1, pp. 66–80, Jul. 2007, doi: 10.1016/j.rse.2006.12.006.

[24]   S. A. Parks, G. K. Dillon, and C. Miller, “A New Metric for Quantifying Burn Severity: The Relativized Burn Ratio,” Remote Sens., vol. 6, no. 3, Art. no. 3, Mar. 2014, doi: 10.3390/rs6031827.

[25]   I. Gitas, G. Mitri, S. Veraverbeke, and A. Polychronaki, “Advances in remote sensing of post-fire vegetation recovery monitoring—a review,” Remote Sens. Biomass-Princ. Appl., vol. 1, p. 334, 2012.

[26]   S. Veraverbeke, I. Gitas, T. Katagis, A. Polychronaki, B. Somers, and R. Goossens, “Assessing post-fire vegetation recovery using red–near infrared vegetation indices: Accounting for background and vegetation variability,” ISPRS J. Photogramm. Remote Sens., vol. 68, pp. 28–39, Mar. 2012, doi: 10.1016/j.isprsjprs.2011.12.007.

[27]   C. E. Soulard, C. M. Albano, M. L. Villarreal, and J. J. Walker, “Continuous 1985–2012 Landsat Monitoring to Assess Fire Effects on Meadows in Yosemite National Park, California,” Remote Sens., vol. 8, no. 5, Art. no. 5, May 2016, doi: 10.3390/rs8050371.

[28]   Q. Qukula, “Knysna fires sparked by lightning strike, not arson, finds expert. Capetalk,” 2017. https://www.capetalk.co.za/articles/269400/knysna-fires-sparked-by-lightning-strike-not-arson-finds-expert (accessed Mar. 05, 2023).

[29]   News24, “Knysna fire led to largest deployment of firefighting resources in SA history - authorities. News24,” News24, 2017. https://www.news24.com/news24/knysna-fire-led-to-largest-deployment-of-firefighting-resources-in-sa-history-authorities-20170620 (accessed Mar. 05, 2023).

[30]   Algerie Presse Service, “Feux de forêt: plus 100.000 hectares ravagés dans 21 wilayas durant l’été 2021,” 2022. https://www.aps.dz/economie/139524-feux-de-foret-plus-100-000-hectares-ravages-dans-21-wilayas-durant-l-ete-2021 (accessed Mar. 05, 2023).

[31]   Copernicus Emergency Management Service, “Algeria Forest Fires. European Commission.,” Copernicus EMS - Mapping, 2021. http://emergency.copernicus.eu/mapping/ems/algeria-forest-fires (accessed Mar. 05, 2023).