Application of Multi-Criteria Decision Making Models in Forest Fire Management


 The study on effective factors of forest fire prevention policy is helpful to reduce forest fire impacts on extensive environmental damage in the long-term period. In other words, forest fire management is the result of a complex interaction among criteria. The present study aims to create a scientific analysis of the most effective criteria based on TOPSIS and SAW methods in the Arasbaran forest. The five top optimal criteria selection by TOPSIS method introduced that “association and cooperation between the executive and responsible institutions” have the first rank (CCi+=0.85), “Lack of deterrence law in dealing with forest fire offenders in human–caused forest fires” has the second rank (CCi+= 0.84) and followed by “Lack of up-to-date scientific information on susceptible areas in the region”, “Increasing the cooperation of NGOs and increase public trust”; and “Lack of forest road network access to ignite regions” (CCi+= 0.789; 0.787; 0.77, respectively). The five top optimal criteria resulting from the SAW method showed that “Local people participations” provide the highest score (FS=0.39) and followed by “association and cooperation between the executive and responsible institutions (FS=0.39), “Increasing the cooperation of non-governmental organizations (NGOs) and increase public trust” (FS=0.36), “Raising awareness of the position of natural resources among local peoples and attracting their cooperation” (FS=0.35) and “Optimal Use past experiences” (FS=0.34). It is suggested that evaluating the ecological and environmental factors affecting the forest fire occurrence and extension could become a set of complement factors to setting management criteria for demonstrating the best management strategies.


Introduction
Forests cover about one-third of the earth surface and considered as fundamental natural resources play an essential role in purifying the air and reduction of greenhouse gases effects, carbon sequestration, preventing flood, protect soil and water erosions, and many other advantages to help move fast towards forest sustainable development (Zandebasiri and Pourhashemi, 2016). However, forest fires as complex a natural phenomenon are those which burn forest vegetative cover that operated by various natural and human factors, it can become a threat to the forest ecological and economical services with some potentially negative impacts on ecosystems. A forest fire can cause serious destructing in forest composition, biodiversity, and structure (Sharma et al., 2012;Carvalho et al., 2019). In addition, Forest fires are known as one of the major causes of ecological disturbance and environmental concerns especially in mountainous deciduous forests (Vadrevu et al., 2010). Iran has witnessed many forest fire events in the past to the present and is a susceptible region of this disaster and forest fires became a significant problem in Iran nowadays According to the Statistics Center of Iran, 610 fires occurred in the forests of Iran and destroyed an area of 5,694 ha from 2000 to 2012. Proper conservation strategies for preventing forest fire occurrences are an effective way of reducing the damages of this natural hazard (Moayedi et al. 2020). Forest fires play an influential role in forest ecosystem services. Therefore, it is essential to predict the forest fires detrimental effects on natural resources and ecosystem services (Vila-Vilardell et al., 2020). Researches have recognized that forest fires are enhanced by changes in climates such as longer dry seasons, rainfall reduction, and frequent extreme temperatures (Pourghasemi et al., 2020). Therefore, increasing the frequency and severity of wildfires is a major problem in certain regions of Iran. Forest fires have become an important issue in recent years. Many studies have been conducted to forest fire susceptibility analysis using various kinds of statistical models and various researches are being done in different parts of the world on different aspects of this issue. Pourghasemi et al. (2020) assessed the forest fire influences and mapping in Fars province in Iran by using machine learning techniques and spatial modeling. Yang et al., (2019) proposed a demand forecasting model based on Index Fuzzy Segmentation (IFS) and TOPSIS in order to effectively predict the number of firefighting helicopters needed in forest fires. Results demonstrated that the demand forecasting model based on these two methods have strong feasibility and rationality, which provides a scientific method for predicting the number of forest firefighting helicopter demand resources. Ghazanfar Pour et al., (2017) identified the most important effective factors of the forest fire control at Golestan Forest in the north of Iran by SWAT and AHP models. The results showed that the measure of solidarity between organizations, the availability of the different parts of the forest, and constructing the forest road network were the most vital factors in forest fire management. Gungoroglu (2017) determined four forest fire risk criteria consisting of socioeconomic, topographic, climatic, and stand structure by using the Fuzzy AHP method and produced the risk map in order to low and high risks levels in Turkey. Zandebasiri and Pourhashemi (2016) examined the strengths and weaknesses of some of the MCDM methods consisting of AHP, FAHP, ANP, TOPSIS, VIKOR, WSM, DEA, Voting methods, PROMETHEE, and ELECTRE to choose an optimal one for decision making in forest management. AHP and SWAT methods were preferred among all methods in forest management. Sharma et al., (2012) analyzed the knowledge-based information to make strategies for forest fire management by using a Fuzzy-AHP approach in India. Findings were including the ranges of low to high forest fire risk zones according to environmental features such as topography, climate conditions, slope, aspect and etc. The present research aims at analyzing the comparison of two multi-criteria decision-making methods included Simple Additive Weighting (SAW) and Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) for determining the most important forest fire prevention factors and management strategies focusing on the Arasbaran Forest region. This study is a novelty study in this case in this region.

Methods and Materials
-Description of the study area Arasbaran forest (160000 ha) is situated between longitude 46° 39 50 and 47° 1 48 E and latitude 38° 43 41 to 39° 8 11 N in the East Azarbayjan province, Northwest of Iran. The location of the mountainous study area is shown in Figure 1. The main tree species is broad-leaved consisted of Quercus macranthera, Quercus petraea, Carpinus betulus, Acer compestre, Acer monspessulanum, and many shrubs species. Taxus baccata and Juniperus foetidissima are the most important conifers of this region. The average total annual temperature is 2-17 C and the average total annual precipitation is 300-600 mm. This area is characterized by special climatic features,  Table  2 for evaluating the importance of each criterion. 15 experts who had an average of 10 years of work experience in the forestry department and more than half of the members had a higher education than a master's degree in Natural Resource Management and were specialists in Forestry and Rangeland and Watershed Management filled out the questionnaires according to their expert opinion. The expert team was asked to determine the value of 1 to 9 for selecting the more effective criteria. The higher numerical value indicated more preference for a criterion. Multi-criteria decision making (MCDM) techniques are useful tools to help decision-makers to select options in the case of discrete problems that refers to making the choice of the best alternative from among a limited set of decision alternatives in terms of multiple, usually conflicting criteria. In other words, the MCDM technique is based on obtaining the alternative that approaches the most ideal alternative (Roszkowska, 2011). Among many multi-criteria techniques TOPSIS and SAW as the most frequently used methods were selected in this study.
-Technique for Order Performance by Similarity to Ideal Solution (TOPSIS) TOPSIS is one of the widely-used classical multi-criteria decision-making methods that was first developed in 1981 by Hwang and Yoon (Hwang and Yoon, 1981). According to the concept of this method, the alternatives are sorted according to their distance from ideal (positive) and inappropriate (negative) solutions at the beginning. Then, the best alternative should have the shortest distance from the positive ideal solution and the farthest from the negative ideal solution

Results
The results of optimal criteria selected by the TOPSIS method showed that C1 which was "association and cooperation between the executive and responsible institutions" has the first rank (CCi + =0.85). C13 which was "Lack of deterrence law in dealing with forest fire offenders in human-caused forest fires" has the second rank (CCi + =0.84) and followed by C9 (Lack of up-todate scientific information on susceptible areas in the region such as forecasting maps, determining the amount of damage to the forest, etc.; CCi + =0.789), C23 (Increasing the cooperation of nongovernmental organizations (NGOs) and increase public trust; CCi + =0.787) and C14 (Lack of forest road network to access the ignite regions; CCi + =0.77). Therefore, there were recognized as the first top-five criteria. In addition, C17 (Construction of stations for measuring effective environmental factors such as anemometer station, temperature recording, etc. in susceptible areas inside the forest; CCi + =0.17), C19 (Upgrading the wireless and fire alarm networks; CCi + =0.12) and C2 (Cooperation and communication of neighboring provinces; CCi + =0.10) were introduced as the least effective criteria in forest fire prevention strategies by TOPSIS method, respectively (Figures 3, 4 and 5).  C1  C2  C3  C4  C5  C6  C7  C8  C9  C10  C11  C12  C13  C14  C15  C16  C17  C18  C19  C20  C21  C22  C23  C24  C25  C26  C27  C28 C1  C2  C3  C4  C5  C6  C7  C8  C9  C10  C11  C12  C13  C14  C15  C16  C17  C18  C19  C20  C21  C22  C23  C24  C25  C26  C27 C28 C29

Criteria
(Cooperation and communication of neighboring provinces; FS=0.10) were the least effectiveness criteria in this method ( Figure 6).

Discussion and Conclusion
Expert knowledge is a basic source for making effective management decisions. There are many MCDM methods for determining the reasonable factors for management in different sectors of forest (Sharma et al., 2012). According to the results of this study, we determined the top main forest fire management strategies based on Expert's knowledge; and Association and cooperation between the executive and responsible institutions (C1), Local people participations (C3), Lack of up-to-date scientific information on susceptible areas in the region such as forecasting maps, determining the amount of damage to the forest, etc. (C9), Lack of deterrence law in dealing with forest fire offenders in human-caused forest fires (C13) and Increasing the cooperation of nongovernmental organizations (NGOs) and increase public trust (C23) were identified and presented by using of TOPSIS and SAW models as two important decision-making methods.
Our results were then compared to those obtained in other areas of the world and comparing these results with other studies displayed that study on the main factors to manage the fires in exotic species plantations of Zimbabwe by Jimu and Nyakudya (2018) supported our results. they also displayed the need of cooperation among government and timber estate owners and community leaders; In addition, perform forest policies, raise education and fire awareness campaigns, strengthen the linkage between indigenous communities and government, managing the maintain fire prevention were introduced as the main strategies in that study area. According to the reports of Eugenio et al. (2019), meteorological variables such as precipitation and air temperature are directly and indirectly correlated with the occurrence, propagation and distribution of wildfires. According to their results, fire hazard zonation is an important factor to manage the beginning, the duration and the end of fire hazard. Moayedi et al. (2020) proposed that providing a high-quality forest fire susceptibility map is an important task for fire risk management. They emphasized on the criterion of the obtain susceptibility analysis of natural hazards and prepare maps as a guide in the risk management and prevention of the fire and fundamental prerequisite for future planning  The study's result showed the regions with a high, moderate, and low susceptible areas to forest fire with knowledgebased factors. They introduced that emergency management plays an important role in the prevention, control and management of disasters within a quick time period. We used this criterion as Criterion 9 in our study that ranked third in the TOPSIS method. Nilsson et al. (2016) presented the combination of two MCDM methods, AHP and TOPSIS to show that this combination was easy to implement in participatory forest planning to create a wide array of management plans. Therefore, the combined MCDM approach can be used for ranking a set of long-term management plans with consideration to multiple objectives. A comparison of the results of ranking in our study revealed that two MCDM methods had similar results so that Criterion 1 and 23 were identified as the best and Criterion 2 as the weakest effective criteria in fire management in both methods. Therefore, we introduce that the combination of TOPSIS and SAW methods can be effective for selecting and ranking the most effective forest fire management strategies. Zandebasiri and Pourhashemi (2016) analyzed the most important MCDM method's strengths and weaknesses in forest management and concluded that the TOPSIS method was the optimal method based on its highest score in evaluation "accuracy of results ". It seems due to defining positive and negative ideal options that cause the accuracy of the method has the highest score among others. The TOPSIS method can evaluate a large number of alternatives and is relatively easy to implement but, the low sensitivity of analysis and team decision making were introduced the disadvantages of this method ( On aggregate, with the accomplishment of the present study, forest fire management and firefighting are influenced by various criteria. The results of this study, based on the experts' knowledge of forest firefighting who had much experience in this field and methods of MCDM (TOPSIS and SAW) showed that association and cooperation between the executive and responsible institutions, increasing the cooperation of non-governmental organizations (NGOs), and increase public trust were the most effective factors and communication of neighboring provinces were the least effective criteria in fire management at the Arasbaran forest region.

Declarations:
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable. Table 2. Questionnaire The importance and impact of the following criteria on fire crisis management in Arasbaran forests Please enter the numbers 1 (least important) to 9 (most important) in front of each factor according to your expert opinion: Profile of respondents: Years of work experience: Education: Specialization: Impact s Criteria 1 2 3 4 5 6 7 8 9 Positiv e C1 Association and cooperation between the executive and responsible institutions C2 Cooperation and communication of neighboring provinces C3 Local people participations C4 Optimal Use of past experiences C5 Allocating additional funding as appropriate C6 Providing detailed management plans Negati ve C7 Lack of firefighters and inadequate implementation of prevention and fire extinguishing operations. C8 Poor forest monitoring, especially during peak fire times C9 Lack of up-to-date scientific information on susceptible areas in the region such as forecasting maps, determining the amount of damage to the forest, etc. C1 0 Lack of dedicated firefighting equipment (such as clothing and portable tools) and high costs of providing other advanced equipment for the organization (such as helicopters) C1 1 Lack of natural or man-made ponds for water storage