A Hybrid Approach Using Fuzzy AHP-TOPSIS Assessing Environmental Conflicts in the Titan Mining Industry along Central Coast Vietnam

Environmental conflict management gains significance in rational use of natural resources, ecosystem preservation and environmental planning for mineral mines. In Central Coast Vietnam, titan mines are subject to conflicting use and management decisions. The paper deals with an empirical research on applying a combination of the fuzzy Analytic Hierarchy Process (AHP) and the fuzzy Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) to measure environmental conflicts emerging as a result of titan mining in Vietnam. The methodology used in the paper combines the fuzzy AHP and the fuzzy TOPSIS to rank environmental conflicts and propose conflict prevention solutions in the titan mining industry of Ky Khang coastal commune (Ky Anh district, Central Coast Vietnam). Data was collected by using a questionnaire with 15 locals, 8 communal authorities, 2 district authorities, and 12 scientific experts on titan mining, environmental geology, and sustainability management. The result shows that, titan mining conflicts with the eight criteria of economic sectors at five alternative sites including beach, protected forest, agricultural area, settlement area, and industrial area. The conflicts between titan mining and forestry, agriculture, settlements, fishing and aquaculture are highly valued. The beach area shows most environmental conflict as a result of titan mining, followed by the agricultural area and settlement area. Based on the empirical findings, legal and procedural tools such as environmental impact assessments, strategic environmental assessments, integrated coastal zone management, marine spatial planning, and multi-planning integration advancing environmental management for titan mines in Vietnam are suggested.


Introduction
Environmental conflicts, which originate as a result of environmental pollutions, resource use competition, and social conflicts, emerge when stakeholders take part in activities with contradictory interests, values, power, perceptions, and goals. Environmental conflicts cover different issues: Biodiversity conflicts [1][2][3], coastal zone conflicts [4][5][6][7], air pollution conflicts [8], land use conflicts [9,10], and water conflicts [11][12][13][14]. Most recently, environmental conflict is considered in relation to economic, since 1997 [33]. Ky Khang borders the Vung Ang Economic Zone (EZ), which is one of the seven key coastal economic zones prioritized by the Vietnamese Government for the period 2016-2020. Over 350 enterprises operate in the Vung Ang EZ, targeting steel smelting, thermal power plants, electricity generation, and deep-water port services. This industrial package puts fishermen's livelihood and sea water quality under pressure. An environmental incident rises from Vung Ang EZ in 2016, attracting both domestic and international attention, is the "Formosa environmental disaster". This problem came from a large source of toxic waste produced by the Formosa Ha Tinh Steel Corporation. The disaster, which spread over the coast of Central Vietnam including the Ha Tinh, Quang Binh, Quang Tri and Thua Thien-Hue provinces, killed sea fish, shrimp, clam, and coral reefs [36,37]. Moreover, in the Ky Anh district, Ky Khang is the commune most seriously damaged by natural disasters such as tropical storms and coastline erosion [35]. Combined natural disasters, environmental disasters and environmental pollutions raise environmental conflicts between the titan mining industry and the other economic sectors (agriculture, aquaculture, fishing, tourism, and forestry).
Appl. Sci. 2019, 9, x FOR PEER REVIEW 3 of 14 2020. Over 350 enterprises operate in the Vung Ang EZ, targeting steel smelting, thermal power plants, electricity generation, and deep-water port services. This industrial package puts fishermen's livelihood and sea water quality under pressure. An environmental incident rises from Vung Ang EZ in 2016, attracting both domestic and international attention, is the "Formosa environmental disaster". This problem came from a large source of toxic waste produced by the Formosa Ha Tinh Steel Corporation. The disaster, which spread over the coast of Central Vietnam including the Ha Tinh, Quang Binh, Quang Tri and Thua Thien-Hue provinces, killed sea fish, shrimp, clam, and coral reefs [36,37]. Moreover, in the Ky Anh district, Ky Khang is the commune most seriously damaged by natural disasters such as tropical storms and coastline erosion [35]. Combined natural disasters, environmental disasters and environmental pollutions raise environmental conflicts between the titan mining industry and the other economic sectors (agriculture, aquaculture, fishing, tourism, and forestry).

Fuzzy AHP
The AHP is one of the most common Multi-Criteria Decision Making (MCDM) instruments to deal with quantifiable and intangible criteria, which reflect the relative importance of the alternatives based on constructing a pair-wise comparison matrix [38][39][40]. Fuzzy AHP was developed to determine the weights of multiple criteria [41]. While the traditional AHP faces limitations on information imprecision and vagueness for decision making, the fuzzy AHP solves these problems of imprecision using linguistic variables, which are used to represent the relative importance between each pair of criteria [42]. Five linguistic variables corresponding with their fuzzy numbers are presented in Table 1. The process of Fuzzy AHP is structured in four steps [41]. Let X = {x1, x2…xn} be an object set and G = {g1, g2…gm} be a goal set. According to the extent analysis method [42], each object xi is evaluated

Fuzzy AHP
The AHP is one of the most common Multi-Criteria Decision Making (MCDM) instruments to deal with quantifiable and intangible criteria, which reflect the relative importance of the alternatives based on constructing a pair-wise comparison matrix [38][39][40]. Fuzzy AHP was developed to determine the weights of multiple criteria [41]. While the traditional AHP faces limitations on information imprecision and vagueness for decision making, the fuzzy AHP solves these problems of imprecision using linguistic variables, which are used to represent the relative importance between each pair of criteria [42]. Five linguistic variables corresponding with their fuzzy numbers are presented in Table 1. Table 1. Linguistic variables and fuzzy numbers [42].

Linguistic Variables Fuzzy Numbers
Extreme importance (EXI) (7; 9; 9) Very strong importance (VSI) ( The process of Fuzzy AHP is structured in four steps [41]. Let X = {x 1 , x 2 . . . x n } be an object set and G = {g 1 , g 2 . . . g m } be a goal set. According to the extent analysis method [42], each object x i is evaluated by performing an extent analysis for each goal, g i . Therefore, m extent analysis values for each object can be obtained by using following notation: M gi 1 , M gi 2 ...M gi m . Step 1. The fuzzy synthetic extent value (S i ) with respect to the ith criterion is calculated in the following way: With: where: l is the lower limit value; m is the most promising value; u is the upper limit value.
Step 2. The degree of possibility of S 2 (l 2 , m 2 , u 2 ) ≥ S 1 (l 1 , m 1 , u 1 ) is defined as: where: x and y are the values on the axis of membership function of each criterion.
Step 3. The possibility that a convex fuzzy number S is greater than k convex fuzzy numbers S i i = 1, k is defined as: Step 4. Calculate the normalized weight vectors W' where: d (A i ) = minV(S i ≥ S t ); i = 1, n; t = 1, n; and i t The normalized weight vectors W' is generated according to the pairwise comparisons of the criteria of the involved respondents. As far as the important the corresponding criterion is concerned, the higher the weight, the more important the corresponding criterion.

Fuzzy TOPSIS
The fuzzy TOPSIS was developed based on the attribute of the shortest and the longest distance from the positive ideal solution and the negative ideal solution [43]. The best alternative has the shortest distance to the positive ideal solution, and the longest distance from the negative ideal solution. The classical TOPSIS assumes that individual preferences are assigned with crisp values. However, one should consider uncertainty and imprecision in the environmental practices. Fuzzy TOPSIS is more feasible because it incorporates the fuzzy environment uncertainty in decision making.
The process of Fuzzy TOPSIS used in this study follows six steps [43].
Step 1. Calculate the aggregate fuzzy ratings for the solutions.
If x ijt = e ijt , f ijt , g ijt , i = 1, n, j = 1, m, t = 1, l is the fuzzy aggregated rating of solution A i , by decision maker D t , with respect to each criteria C j . The fuzzy aggregated rating x ij = e ij , f ij , g ij , is given by: where: e ij = 1 l l t=1 e ijt , f ij = 1 l l t=1 f ijt , and g ij = 1 l l t=1 g ijt Step 2. Calculate normalized fuzzy decision matrix.
Data is normalized to obtain a comparable scale using linear scale transformation. Suppose r ij = a ij , b ij , c ij is the mean of alternative solutions i for criterion j. The normalized value x ij can be calculated as: The weighted-normalized value G i is given by multiplying the normalized value x ij of the decision matrix by the weight assigned to the criterion j.
Step 4. Determine the fuzzy positive ideal and negative ideal solutions.
The positive ideal solution maximizes the benefit criteria and minimizes the cost criteria; whereas the negative ideal solution maximizes the cost criteria and minimizes the benefit criteria. The positive and negative ideal solutions are found out in the following way: Calculate the distance/separation from: * Positive Ideal Separation (d + ): where: d + i and d − i are the distances of each alternative A i from positive and negative ideal solutions.
Step 5. Calculate the Relative closeness coefficient (CC) to the ideal solution.
Step 6. Rank alternatives The rank of considered alternatives can be decided, according to the descending order of closeness coefficient (CC). When the closeness coefficient is closer to 1 the corresponding solution is the optimal one.

Data Collection
The hybrid approach using Fuzzy AHP-TOPSIS provides a quantitative logical and systematic framework to identify critical issues, attach relative priorities to those issues, choose best compromise alternatives, and facilitate communication towards general acceptance [29,44]. This paper considered 8 criteria, representing the main economic sectors of Ky Khang, and 5 different alternative sites, representing conflict hotspots. Fuzzy AHP is combined with Fuzzy TOPSIS to estimate weights of the criteria and to prioritize alternative sites according to the intensity of environmental conflicts. Data on weighting sector criteria and prioritizing alternatives sites in the Ky Khang commune are collected using a questionnaire with 15 locals, 8 communal authorities, 2 district authorities, and 12 scientific experts on titan mining, environmental geology, and sustainability. Data were collected during a field trip in March 2018. All respondents inhabit in the study area (locals and authorities) or are knowledgeable about the study area and about scientific problems related to titan mining. The questionnaire allows inventory the opinion of the respondents on the pair wise comparison matrix by using Likert 5 scale, indicating five pre-coded responses with the neutral point (point 3) being neither agree nor disagree. Respondents took about 30 min to complete the questionnaire.

Determining Criteria of Sectors and Alternative Sites
In the considered case study, in the area of Ky Khang, couples of environmental conflicts between the titan mining industry and other economic sectors are found across conflict hotspots. These couples are indicated by criteria in Fuzzy AHP model. Conflict hotspots are defined by alternative sites in Fuzzy TOPSIS model. It is supposed that conflicts occur based on eight criteria (C j ) and are responsible for five alternatives sites (A i ). The eight criteria entail agriculture (C1), aquaculture (C2), fishing (C3), salt production (C4), tourism (C5), forestry (C6), settlement (C7), and industry (C8) (Figure 2). Five alternatives sites include the beach (A1), a protected forest area (A2), an agricultural area (A3), a settlement area (A4), and an industrial area (A5). Since titan mining is popular along the coast, it conflicts with the eight criteria of sectors at the five alternatives sites. Impacts concern land use and competition with other sectors, use of fresh water during mining process, and emission of pollutants to air, soil and sediments. Consequently, mining destroys protection forests near the coast, reduces both surface and underground water quality and quantity, increases salinity, and destroys beaches.
In beach area (A1), titan mining impacts aquaculture, fishing, and tourism negatively by destroying beaches, eroding the coastline, and changing landscapes visually. Parts of titanium dumps were filled by clastic sedimentary rocks. Sand dunes have changed leaving hollow pits and deep holes. New sandy dunes of 6-10 m have appeared, consisting of silky sand, and being mobile by the wind. Mining leaves disposal sites and reservoirs over large areas.
Building infrastructure in a protected forest (A2) needs cutting trees, takes aquaculture ponds and mangrove farms, and kills sea fishes as a results of environmental pollution.
Titan mining takes arable land and decreases crop yields in agricultural area (A3). It causes competition spatially between titan mining and areas for salt production and decreases the quality and the yields of the salt due to pollutants. Crops on the fields and houses are flooded by sand. Mining damages soil surface: A large amount of soil and rocks are removed from the ground and leave holes behind. Surface water is polluted by overflowing acid effects, toxic pollutants and water-soluble solids release from deposits and sediments. Mine waste discharges sediments in rivers and streams.
Also, titan mining affects negatively buildings, heritage sites, and tourism infrastructures in the settlement area (A4). In industrial area (A5), titan mining and other industrial activities emit pollutants in the environment, which stress both locals and tourists. Mining machines, pumps and transport cause noise, which directly affects residents and tourists.

Weighting Criteria
The opinions of the different involved respondents (37 in total), collected by means of a questionnaire, were used to generate the pairwise comparisons of the criteria and determine the final importance levels represented in the decision matrices. The respondents determine the level of environmental conflicts between the titan mining industry with other economic sectors for five alternative sites. The evaluation of 5 decision matrices on 8 criteria is presented in Table 2. For example, comparing C1 and C2, C1 is considered of "very strong importance" compared to C2 for DM1 (value in the 4th column and in the 2nd raw). Consequently, the reciprocal value is assigned when C2 is compared to C1 for the same DM (3rd column, 7th raw).
Equations (1) and (2) are used to calculate the levels of the comparison between two fuzzy numbers based on the decision matrix. The relationship between a fuzzy number which is higher than the remaining fuzzy numbers, is calculated using Equations (3) and (4). Table 3 shows the results of priority weighting of the criteria based on the Equation (5). Forestry shows the highest score (0.131), which indicates titan mining has the most significant impacts on the forest plantations. Also, the conflicts between titan mining and agriculture, settlements, fishing and aquaculture are highly valued: agriculture shows the second weight score (0.129), followed by settlements (0.128), fishing (0.127), and aquaculture (0.125). Where: EXI = extreme importance; VSI = very strong importance; SI = strong importance; MI = moderate importance; EI = equal importance; the terms in brackets "()" indicate the reciprocal of them. Table 3. Priority weighting matrix.   Table 4). Formula (6) allows evaluating the five decision matrices on the level of conflict. Formulas (7) and (8) are used to standardize the criteria. Formula (9) allows determining the value of the ratio weighted normalized and the distances of each alternative to the positive and negative ideal points. The values of separation from positive solution (d + ) and Separation from negative solution (d − ) are calculated by using Formulas (10)- (13). Values of closeness coefficient (CC) are calculated from values of d + and d − by means of formula (14). The weighted normalized decision matrices indicate the highest and the lowest weighted values belong to forestry and tourism respectively (w (C6) = 0.131; w (C5) = 0.116) ( Table 5).   The ideal solution comprises all of best values possible of criteria, whereas the negative ideal solution consists of all worst value possible of criteria. Calculated separation from negative ideal solution (d − ) indicates beach area (A1), agricultural area (A3) and settlement area (A4) have the highest scores (0.56, 0.46, and 0.44 respectively); whereas they get lowest scores for separation from positive ideal solution (d + ) (7.47, 7.58, and 7.60 respectively). According to closeness coefficient, the beach area (A1) shows most environmental conflict as a result of titan mining (CC (A1) = 0.0703), followed by the agricultural area (CC (A3) = 0.0575), and settlement area (CC (A4) = 0.0549) ( Table 6). Since the worst alternative sites have farthest distance from the ideal solution and the shortest distance from the negative ideal solution, it is necessary to pay more attention on environmental conflicts in these alternative sites.

Conclusions
The paper presents empirical research on environmental conflicts emerging as a result of recent development of titan mining industry along the Central Coast Vietnam. The considered Ky Khang coastal area is located in an economically fast developing area. The resources are subject to conflicting use and management decisions. Coastal zone conflicts are fueled by a combination of other conflicts, which demonstrates the unique character of environmental conflicts along coast. Titan mining is a major problem resulting in most serious impacts on the ecosystems and the environment along coasts. Mining destroys protected forest areas, erodes coastlines, affects the ground water level, increases salinity and modifies natural landscapes. The risk of environmental conflicts between socio-economic groups along the coast increases during the mineral exploitation period.
A combination of Fuzzy AHP and Fuzzy TOPSIS shows its advantages in connecting decision makers with conflicting objectives to reach consensus. The systematic and logical approach is a strength since it allows solving complex multi-person and multi-criteria decision problems by weighting environmental conflicts and relating them to alternative sites. The model contributes to the group decision-making process taking into account the affected sites surrounding the titan mines. Determining the intensity of the conflicts and the alternative sites in which titan mining occurs is imperative. According to the results of this study, titan mining affects most seriously the forests. Since only few forests are left along the coast, environmental conflicts are obvious for protected forests. The raising of beach tourism, agricultural production, and housing all demand more space and compete with titan mining. Therefore, it is necessary to pay more attention to solutions mitigating conflicts in these sites.
To the best of our knowledge, till now no previous study has identified the ranks of environmental conflicts and the priorities of conflict prevention solutions in the case of coastal areas of Vietnam. Therefore, the paper makes an attempt to discuss management implications for environmental conflicts in mineral mines for Vietnam. These environmental conflicts are at risk of escalation and intensification in case they are not managed well [44][45][46]. The findings of this study suggested decisions should use decision science in identifying priority of prevention solutions for each environmental conflict hotspot. In the titan mine of studied Ky Khang, key issues of environmental conflicts concern climate change, biodiversity, environmental air quality, forestry, fresh water, and land resources. Environmental conflicts raise in Ky Khang and in the Vung Ang EZ because of the limited land areas and freshwater resources. Surface water is salinized by the mine activities and useless for agri-and aquaculture. Air is polluted by smoke, toxic gases and dust from the mining activities and electricity generation. While agriculture, forestry, aquaculture, salt production, tourism, heritage preservation, and human health all are affected by titan mining, the couple of conflicts titan mining-forestry should be considered most carefully in the study area. Managing environmental conflicts pertains to the available tools and approaches, and recommendations arising from these study results provide a scientific basis for policy-makers to mitigate the negative impacts of the conflicts. Since environmental conflicts dovetail in political, economic, social and ecological contexts, their management should focus on integrated measures [47,48]. Environmental impact assessments (EIA) and strategic environmental assessments (SEA) are important legal and procedural tools to manage the environment, in particular with regard to the current and intended development [49,50]. Moreover, integrated coastal zone management (ICZM), marine spatial planning (MSP), and multi-planning integration advancing coastal zone management are favorable to titan mines along coasts. In reality, there is lack of planning for natural resource use in the Central Coast Vietnam [35,51]. Consequently, resources in this area are depleted and risks of conflicts between economic groups using these natural resources are common. Over-all, conflicts will likely become more severe and serious as long as resources decline both in quality and quantity and environmental pollution increases. Developing natural resources and environmental management strategies and models of sustainable consumption of resources according to a sustainable development strategy are required. Potential conflicts are identified, and preventive measures are proposed.