KOŠICKÁ BEZPEČNOSTNÁ REVUE KOSICE SECURITY REVUE Vol. 8, No. 1 (2018), p. 35 44 ISSN 1338-4880 (print ver.), ISSN 1338-6956 (online ver.) Study of a forest fire behaviour in changed wind speed conditions Štúdium správania sa lesného požiaru v podmienkach so zmenenou rýchlosťou vetra Andrea MAJLINGOVÁ 1, Miroslava SMOLÁROVÁ 2 and Rastislav VEĽAS 3 1,2,3 Technical University in Zvolen, Slovakia The manuscript was received on 13. 04. 2018 and was accepted after revision for publication on 01. 06. 2018. Abstract: Forest fires represent threat mostly in countries situated in temperate climate zone, however, because of the changing climate, longer periods of high temperatures and drought, we can expect that their number and extent will increase in Slovakia, too. There are several factors influencing the initiation and further propagation of forest fire. Among the most important factors belong the volume and type of fuel available, topography, meteorological conditions and the time of fire. Those influence mostly the fire propagation, i.e. fire behaviour. In the paper, we introduce the results of a study focusing the fire behaviour in changing wind speed conditions. To model the fire behaviour, the FARSITE software was applied. In the modelling, we applied the data on real forest fire that occurred in the Slovensky Raj National Park, Koc locality in 2007. We modelled the fire behaviour for 24 h, using the specific wind speeds (0 km/h, 34 km/h, 68 km/h). According to the modelling results, we analysed the trend of fire propagation for each wind speed and provided the correlation analysis between the fire site extent (fire area and fire perimeter) and wind speed in Statistica 12 software. The fire propagation in time trend analysis results showed, that during the night, the extent of fire site was not increasing, probably because of the lower night temperatures and increasing moisture of the vegetation, over the moisture of extinction critical point, but not because of wind. The process of fire propagation started in the morning, after sun rise, again. The correlation analysis results showed very strong dependency between the fire site extent and wind speed. The results of this study have its immediate utilisation in the fire-fighters practice as for intervention commanders decision support as for fire-fighters education and training. Keywords: dependency, FARSITE, fire behaviour, forest fire, wind speed This work is licensed under a Creative Commons Attribution 4.0 International License. 35
Abstrakt: Lesné požiare predstavujú hrozbu najmä pre krajiny situované v teplej klimatickej oblasti, avšak, z dôvodu zmeny klímy sa dá očakávať nárast ich počtu a rozlohy aj na Slovensku. Vo všeobecnosti existuje niekoľko faktorov, ktoré majú vplyv na vznik a následné šírenie požiaru. Medzi najvýznamnejšie patria množstvo a typ dostupného paliva, topografia, meteorologická situácia a čas vzniku a trvania požiaru. Tieto ovplyvňujú najmä šírenie, t. j. samotné správanie sa požiaru. V článku predstavujeme výsledky štúdie zameranej na posúdenie správania sa požiaru v podmienkach meniacej sa rýchlosti vetra. Na účely modelovania bol použitý softvér FARSITE. V procese modelovania sme použili údaje z reálneho požiaru, ktorý sa vyskytol v roku 2007 na území Národného parku Slovenský raj, konkrétne v lokalite Koč. Požiar sme modelovali počas 24 h, s využitím konkrétnych rýchlostí vetra (0 km/h, 34 km/h, 68 km/h). Vychádzajúc z výsledkov modelovania sme analyzovali trend šírenia požiaru pre každú z modelovaných rýchlostí vetra a programe Statistica 12 vykonali aj korelačnú analýzu medzi veľkosťou požiariska a rýchlosťou vetra. Výsledky analýzy trendu šírenia požiaru v čase poukázali na fakt, že počas večera a noci sa veľkosť požiariska nezväčšovala, a to pravdepodobne z dôvodu nižších nočných teplôt a stúpajúcej vlhkosti materiálu, pravdepodobne až nad hranicu vlhkosti vyhasnutia, avšak nebolo to kvôli vetru. Šírenie požiaru nastalo znova až v ranných hodinách, po východe Slnka. Výsledky korelačnej analýzy poukázali na veľmi tesnú závislosť medzi veľkosťou požiariska a rýchlosťou vetra. Výsledky štúdie majú svoje okamžité uplatnenie v hasičskej praxi, a to ako podpora rozhodovania veliteľov zásahu a tiež vo vzdelávaní a výcviku hasičov. Kľúčové slová: závislosť, FARSITE, správanie sa požiaru, rýchlosť vetra Introduction Forest fire is a well-known phenomenon also in Slovakia. There are known totally five types of forest fire in Slovakia, in contrast to three types distinguished in the world. Besides the underground, ground and crown fire, we distinguish also the hollow tree fire and wind disaster disturbed area fire. The most dangerous fire is the crown fire. This type of fire occurs rarely in Slovak conditions. The most often occurring is the ground fire and the underground fire occurring, mostly in chalky areas. Research on forest fire includes more issues in the world: forest fire risk assessment, building forest fire warning systems, fire behaviour modelling, research on impact of individual fuel [4,18], topography [9,21], meteorological and wind condition on fire behaviour [22, 23], implementation of UAV to fire monitoring [16], CCTV fire detection systems implementation in the fire prevention sphere [17, 15]. Research on forest fires in Slovakia focuses mostly the fire risk assessment [19,20], fire behaviour modelling [6,7] and testing the fire properties of woody fuel [5,12,13,14]. There are only few works dealing with assessment of forest fuel volume [10,11] or research on factors influencing the forest fire ignition and propagation [1,2,8]. This work is one of the first studies in Slovakia which deals with investigation of the wind speed change on fire behaviour, expressed by the fire site area and perimeter in this case. Information and knowledge obtaining by tackling this issue is creates a very good base to build the knowledge base on forest fire behaviour in Slovak conditions. This information is necessary not only for fire behaviour modelling itself, but also for 36
explanation of modelling results, for enhancement of decision support of fire incident commanders as well as for education and training of fire-fighters. 1. Forest fires in Slovakia Forest fire in Slovakia occur mostly during the spring and summer season. They often occur as a consequence of human activity in forest, either intended or careless. In Fig. 1 we introduce the trend in forest fire occurrence in forests of the Slovak Republic in period 2008-2017. Fig. 1 Forest fire occurrence in the Slovak Republic in period 2008-2017 The data on forest fire number were provided by the Fire Research Institute of the Ministry of Interior of the Slovak Republic. 2. Forest fire behaviour principles Forest fire behaviour is in many aspects much more complicated as the other types of fire. The parameters which influence the process of burning are constantly changing because of the change of the conditions in the fire site itself. The change is either advantage or disadvantage for controlling the fire and fire-fighting [3]. The success or failure of fire-fighting activities in connection with the occurrence of a forest fire depends on the ability of the commander to interpret the situation or, in many cases, to predict what will happen when different influences, whether internal or external, interact. The fire-fighting process consists of the same elements as other types of fire: initiator (heat), fuel, oxygen access and time. As with other types of fire, its extinguishing is directly dependent on the quickest and most effective interruption of 37
the burning process. However, in case of forest fires, or wildland fire, apart from the traditional triangle of burning, the so-called "tetrahedron of burning" is used, due to the greater complexity of the burning process [3]. Among the tetrahedron features belong: fuel parameters, meteorological situation, topography and fire time (from its ignition up to the completing the firefighting activities). When the fire is propagating, those factors, which are mutually interconnected are constantly changing. As mentioned above, during a fire, from its ignition up to removing, fire behaviour, within the fire site changes a hundred, perhaps a thousand times. This is especially important from the incident commander point of view, who should understand these changes and their significance for the further propagation of the fire, the further performance of localization and fire-fighting activities in the fire-affected area. The most important weather parameter in terms of fire is the wind. Wind is a direct result of the cooling and heating of the Earth's surface, which creates a varied air pressure. From the point of view of fire-fighting activities during a fire, the wind has either positive or negative effects, either because it slows the fire action or stimulates it. As mentioned above, the basic effects of the wind on a fire include changing the rate of fire propagation and changing the direction of its spread. However, it is also worth mentioning the fact that a double increase in wind speed does not automatically mean a double acceleration of fire propagation. In this case, the fire rate increases in the geometric sequence. In other words, when the fire area is growing rapidly, the impact of the wind on the continuous propagation is multiplied by the power. The prevailing weather conditions are in fact a representation of the set of wind directions and their speeds occurring normally in the given area, even in the time with absence of a fire. What creates them is the uneven heating of the geographical area in combination with the topography of this territory. The results of the long-term observation of fire-fighters during the fighting of wildland fires show that the wind's speed is increasing with the sun rise speeds, while with the sunset its speed gradually decreases. Considering the fact that the surface of the landscape is warming the fastest between six and ninth o`clock in the morning and is cooling at the time of the fifth and seven o`clock in the afternoon, the wind speed is highest right in this time, too. Coleman 3. Experiment The subject of the study was the influence of the wind conditions, especially wind speed, on forest fire initiation and propagation. The object of the study was the territory of Slovensky raj National Park, especially Koc locality, which was affected by forest fire in 2007. The data necessary to describe the position of the fire site and meteorological and wind conditions during the fire were obtained from the District Directorate of Fire and Rescue Service in Presov. 38
The forest fire in Koc locality was observed on 22. 7. 2007 at 2.45 pm. The influence of wind conditions on forest fire behaviour was studied based on the fire site extent (perimeter in km and area in hectares) obtained from modelling, in dependency on changed wind speed. For modelling the forest fire behaviour, we used the FARSITE software. The fire behaviour was modelled from its ignition for a period of 24 h. Modelling results focusing the fire site perimeter and area in studied wind speeds (0, 3, 9, 16, 24, 34, 44, 56, 68 km h -1 ) were further applied in trend analysis focusing the development of fire site extent and in correlation analysis serving for studying the tightness of dependence between the fire site extent and wind speed changed. Correlation analyses were performed in Statistica 12 software. 4. Results and discussion In the analysis of the impact of wind speed on forest fire propagation, we proceeded to focus on analysis of the correlation between the influence of wind speed on the perimeter (km) and the area (ha) of a fire modelled during the first 24 hours of burning. In statistical analyses, however, correlations were calculated only for wind velocities 0, 34, 68 km h -1 and for time steps 1, 3, 5, 10, 15, 20, 24 h from the beginning of the fire. Trend analysis is performed for all twenty-four-time steps, both for the fire site perimeter (Fig. 2) and the fire site area (Fig. 3). Fig. 1 Fire site perimeter development trend in dependency on wind speed From the graph shown in Fig. 2, the gradual increase of the fire site perimeter up to 5 hours from the fire ignition (from 2.45 pm to 7.45 pm) is evident. Subsequently, in the next 12 hours (7.45 pm 6.45 am), the size of the fire site did not change, probably due to lower air temperatures (night) and rising moisture content of forest 39
fuel, probably above the moisture of extinction value. The wind speed did not have a major impact on night-time behaviour, due to the already mentioned rising moisture content of the fuel during the night, which starts to lose moisture after sunrise. There was a gradual increase in the propagation rate of the fire in the morning from 7.45 am, i. e. after 13 hours of fire burning, and it was gradually increasing until 2.45 pm, i.e. 24 hours after fire. This may have been caused by the rising air temperature and the decreasing moisture of the fuel during the day. Based on this, it can be stated that wind has a significant impact on the behaviour of the fire, its propagation, especially during the day. Fig. 3 Fire site area development trend in dependency on wind speed The development of the fire site area depending on the wind speed and the duration of the fire, apparent from the graph in Fig. 3, copies the course of the curves representing the development of the fire site perimeter values over time, see Fig. 2. From the statistical evaluation of the influence of wind speed on the development of the extent of the fire site point of view, basic statistical data and correlation of data defining the extent of the fire site at the specified times and depending on the set wind speeds were also surveyed. Here we present the results (Tab. 1) of the correlation between the wind speed and the fire site perimeter for wind speeds of 0, 34, 68 km h -1 and fire duration time steps of 1, 3, 5, 10, 15, 20, 24 h. 40
Tab. 1 Results of correlation analysis between the wind speed and fire site perimeter Wind speed (km.h -1 ) 0 34 68 Mean Standard Deviation 0 1.000000 0.995526 0.951077 2.93 1.60 34 0.995526 1.000000 0.969772 2.59 1.67 68 0.951077 0.969772 1.000000 3.08 3.08 In Tab. 2, there are introduced the results of correlation analysis between the wind speed and the fire site area. Tab. 2 Correlation analysis between the wind speed and fire site area Wind speed (km.h -1 ) 0 34 68 Mean Standard Deviation 0 1.000000 0.999015 0.986866 52.33 58.11 34 0.999015 1.000000 0.991786 46.67 60.76 68 0.986866 0.991786 1.000000 80.24 132.56 In the tables introduced the correlation coefficients, defining the tightness of the correlation between wind speed and fire site extent, were calculated at a 95% level of significance. From the results, is evident, this is a very tight dependence. Conclusions The results of analyses focusing the effect of the change in wind speed on fire behaviour have not demonstrated the very clear significance of the impact of individual factors on the behaviour of a forest fire that occurred in Slovensky raj, Koc locality, in 2007. The results of the analysis of the trend of fire site extent development revealed interesting facts in the propagation of fire during the night and during the day. They confirm the facts about the behaviour of the forest fires during the day and night that also Coleman [3] described in his book. The results of the correlation analyses showed a very tight dependence of the investigated factors. However, FARSITE modelling results focusing the fire site extent (perimeter and area) did not show the expected logical correctness. In this case, it will be necessary to deals with it more in the future, especially to study the logical links between input fuel parameters and in particular its physical parameters as well as meteorological parameters, not only as defined in FARSITE but in the Rothermel`s model itself, which is a integral part of FARSITE. 41
Acknowledgement This study was granted by the KEGA Grant Agency, project no. KEGA 032PU- 4/2018. References [1] BENEDIK, V. Vplyv vlhkosti podložia, materiálu a sklonu terénu na vznik a rozvoj pozemného lesného požiaru = The influence of the subsoil, material and the slope of the subsoil on the ignition and development of terrestrial wildfire. Ph.D. Dissertation, Technical University in Zvolen, 2014. [2] BENEDIK, V. MARKOVÁ, I. Faktory ovplyvňujúce vznik lesného požiaru = Factors affecting initiation of forest fire. In Požární ochrana 2012 : sborník přednášek XXI. ročníku mezinárodní konference : Ostrava, VŠB-TU 5.-6. září 2012 / ed. Michail Šenovský. Ostrava : Sdružení požárního a bezpečnostního inženýrství, 2012. ISBN 978-80-7385-115-6. ISSN 1803-1803. S. 17-20. [3] COLEMAN, R.J. Management of fire service operations. Massachusetts: Breton Publishers, North Scituate, 1978, 560pp. [4] ESTES, B. L., KNAPP, E. E., SKINNER, C. N., MILLER, J. D., PREISLER, H. K. Factors influencing fire severity under moderate burning conditions in the Klamath Mountains, northern California, USA. Ecosphere, 2017, 8. [5] GALLA, Š., MAJLINGOVÁ, A., SUCHOMEL, J. Fire safety aspects of wood as a renewable energy source. In Advanced materials research. Environmental and safety aspects of renewable materials and energy sources, 2014, Vol. 1001, p. 312-317. [6] GLASA, J. Computer simulation and predicting dangerous forest fire behaviour. In International Journal of Mathematics and Computers in Simulation, 2009, vol. 3, no. 2, pp. 65-72. [7] HALADA, L., WEISENPACHER, P. Principles of Forest Fire Spread Models and their Simulation. In Journal of the Applied Mathematics, Statistics and Informatics, 2005, Vol. 1, No. 1, pp. 3-13. [8] CHROMEK, I., BENEDIK, V., ŠMIGURA, M., HLAVÁČ, P. Ochrana materiálov na báze dreva pred ohňom gélovými prípravkami = Protection of wood based materials against fire by gels. In Acta facultatis xylologiae Zvolen: vedecký časopis Drevárskej fakulty, 2010, roč. 52, č. 2, s. 81-90. [9] JAHDI, R., DARVISHSEFAT, AA, ETEMAD, V., MOSTAFAVI, MA. Wind Effect on Wildfire and Simulation of its Spread (Case Study: Siahkal Forest in Northern Iran). In Journal of Agricultural Science and Technology, 2014, Volume 16, Issue 5, p. 1109-1121. [10] MAJLINGOVÁ, A., SEDLIAK, M., TUČEK, J. Results of forest fuel spatial distribution mapping for fire simulation purposes - case study. In Advanced materials research. Environmental and safety aspects of renewable materials and energy sources, 2014, Vol. 1001, p. 356-361. 42
[11] MAJLINGOVÁ, A., TUČEK, J., VIDA, T. Fuel models specification for fire modeling and simulation purposes using existing information about forests. In Zbornik radova: 10. Medjunarodna konferencija zaštite od požara i eksplozije. Novi Sad, Serbia, 2006, pp. 444 455. [12] MAKOVICKÁ OSVALDOVÁ, L. Vybrané technické parametre ohrozujúce záchranné zložky pri požiaroch dreva = Selected technical parameters putting rescue units in jeopardy in case of wood fires. Habilitation thesis. Zilina: University of Zilina, 2015, 170 p. [13] MITTEROVÁ, I. Evaluation of selected wood species from the fire safety point of view. In Wood and fire safety: 7th international scientific conference: proceedings, ed. Anton Osvald. Ostrava - Zábřeh: Šmíra - Print, 2012, p. 163-170. [14] ORÉMUSOVÁ, E., MAKOVICKÁ OSVALDOVÁ, L., OSVALD, A. Gross calorific value of leaves, bark, and branches of selected deciduous trees. In: Transactions of the VŠB - Technical University of Ostrava: safety engineering series, 2012, Vol. 7, no. 1, p. 32-36. [15] PAJTÍKOVÁ, J., LALKOVIČ, M. Detekčné a monitorovacie protipožiarne systémy význam a použitie =. Prezentácia k Taktickému ukážkovému cvičeniu Lešť 2012. [16] RESTAS, A. Unmanned Aircraft System Applications: Firefighting. Introduction to Unmanned Systems: Air, Ground, See & Space. In: LeMieux, J., Ed., Technologies and Commercial Applications, 2012, LCCN 2012954516. [17] SCHROEDER, D. Evaluation of three wildfire smoke detection systems. In Advantage, 2004, Vol. 5, No. 24, pp. 1 8. [18] TELESCA, L. Fractal solutions for understanding complex systems in earth sciences. Book Series: Springer Earth System Sciences, 2016, p. 117-152, DOI: 10.1007/978-3-319-24675-8_8 [19] TUČEK, J., ŠKVARENINA, J., MINĎÁŠ, J., HOLÉCY, J. Catalogue describing the vulnerability of landscape structures in the Slovak Paradise National Park. In Proceedings of International Workshop on Forest Fire in the Wildland-Urban Interface and Rural Areas in Europe: An Integral Planning and Management Challenge. Athens: Institute of Mediterranean Forest Ecosystems and Forest Products Technology, 2003, p. 73 84. [20] TUČEK, J., MAJLINGOVÁ, A. Forest fire vulnerability analysis. In Bioclimatology and natural hazards. Dordrecht: Springer Science+Business Media B.V., 2009, p. 219-230. [21] SALIS, M., ARCA, B., ALCASENA, F., ARIANOUTSOU, M., BACCIU, V., DUCE, P., DUGUY, B., KOUTSIAS, N., MALLINIS, G., MITSOPOULOS, I. Predicting wildfire spread and behaviour in Mediterranean landscapes. In International Journal of Wildland Fire, 2016, Volume 25, Issue 10, p. Pages: 1015-1032, DOI: 10.1071/WF15081 43
[22] PIMONT, F., DUPUY, J. L., LINN, R. R., PARSONS, R. & MARTIN- STPAUL, N. Representativeness of wind measurements in fire experiments: Lessons learned from large-eddy simulations in a homogeneous forest. Agricultural and Forest Meteorology, 2017, 232, 479-488. [23] VAN BEUSEKOM, A. E., GOULD, W. A., MONMANY, A. C., KHALYANI, A. H., QUINONES, M., FAIN, S. J., ANDRADE-NUNEZ, M. J., GONZALEZ, G. 2018. Fire weather and likelihood: characterizing climate space for fire occurrence and extent in Puerto Rico. Climatic Change, 146, 117-131. Authors: 1 Assoc. prof. Andrea Majlingová, PhD. Technical University in Zvolen, T. G. Masaryka 24, 960 53 Zvolen, Slovakia, e-mail: majlingova@tuzvo.sk 2 Miroslava Smolárová Technical University in Zvolen, T. G. Masaryka 24, 960 53 Zvolen, Slovakia 3 Ing. Rastislav Veľas Technical University in Zvolen, T. G. Masaryka 24, 960 53 Zvolen, Slovakia, e-mail: xvelasr@is.tuzvo.sk 44