International Soil and Water Conservation Research (xxxx) xxxx–xxxx HOSTED BY Contents lists available at ScienceDirect International Soil and Water Conservation Research journal homepage: www.elsevier.com/locate/iswcr Original Research Article Multi-criteria decision analysis for sub-watersheds ranking via the PROMETHEE method☆ ⁎ Tijana Vulević , Nada Dragović University of Belgrade, Faculty of Forestry, Kneza Višeslava 1, 11030 Belgrade, Serbia A R T I C L E I N F O A BS T RAC T Keywords: Sub-watershed Soil erosion Ranking PROMETHEE method Soil and water resources are important elements of the environment that is managed to reduce the erosion rate and the destructive effects of torrential flooding Implementation of the measures to reach this goal requires the ranking of sub-watersheds and areas within the sub-watersheds, from most to least vulnerable, which can be achieved using Multi-criteria decision analysis methods In this paper, using the Preference Ranking Organization Method for Enrichment Evaluations (PROMETHEE) II method ranking of nine sub-watersheds delineated in the Topciderska river watershed, Serbia, was performed by using ArcGIS software The criteria used for determining the order of the most vulnerable sub-watersheds are land cover, rainfall, soil erodibility as well as topography The importance of criteria was determined by using the Analytic Hierarchy Process (AHP) method, and the influence of the criteria weights on ranking results was tested This research identified that the most vulnerable sub-watershed is located in the upper part of the study area, where 15% of the territory is at high risk of erosion Thus, this sub-watershed should have priority for protection through the implementation of appropriate measures and works The results of the PROMETHEE II method and the ArcGIS application represent the valuable information for watershed management planning and implementation of soil erosion and torrent control measures Introduction contribute to decision making even in the case of missing data and lack of field research This approach, known as Multi-attribute decision making or Multi-criteria decision analysis (MCDA) enables decision making when the number of alternatives or actions is evaluated in terms of more, usually conflicting criteria Belton and Stewart (2002) address strengths and weakness of MCDA methods, and classified them in: measurement models (Weighted Sum Method-WSM, Analytic Hierarchy Process-AHP), outranking method (Elimination and Choice Expressing Reality – ELECTRE, Preference Ranking Organization Method -PROMETHEE), and goal, aspiration or reference level models (Goal programming, heuristics, metaheuristics) One of the world-renowned MCDA methods are AHP, ELECTRE and PROMETHEE method widely applied in different field, as well as in the area of soil and water resource protection (such as: Macary, Ombredane & Uny, 2010; Vulević, Dragović, Kostadinov, Simić, & Milovanović, 2015; Krois & Schulte, 2014; Grau et al., 2010) These methods differ in the mathematical foundation and serve to rank actions, to select the most appropriate action, or to sort actions into different categories One way to interpret categories is "ordered and predefined groups" (Ishizaha & Nemery, 2013) that in soil and water Soil erosion by water is a serious environmental problem that causes significant soil loss and increases the risk of flooding when sediment load is transported through the water courses To avoid or mitigate this and numerous other undesirable consequences of soil erosion (e.g siltation of accumulation and water pollution), it is necessary to implement measures and work in the watershed in a predefined order The most erodible part of the watershed (subwatershed) has a priority for human reaction – through reforestation, technical object construction, and other measures Determining the sub-watersheds of the highest priority for protection is important due to financial and human resources planning, using soil erosion estimation and modeling, applying empirical, conceptual or physical models (see Merritt, Letcher, & Jakeman, 2003) Some of these models are appropriate for specific erosion process, requiring differential data which is difficult to obtain or are too complex to implement (Fernández & Vega, 2016; Wang et al., 2016) Beside this, there is an approach based on the use of quantitative and quantitative data and expert judgment and knowledge which can Peer review under responsibility of International Research and Training Center on Erosion and Sedimentation and China Water and Power Press ⁎ Corresponding author E-mail address: tijana.andrijanic@sfb.bg.ac.rs (T Vulević) http://dx.doi.org/10.1016/j.iswcr.2017.01.003 Received 16 August 2016; Received in revised form 26 January 2017; Accepted 27 January 2017 2095-6339/ © 2017 International Research and Training Center on Erosion and Sedimentation and China Water and Power Press Production and hosting by Elsevier B.V This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/) Please cite this article as: Vulevic, T., International Soil and Water Conservation Research (2017), http://dx.doi.org/10.1016/j.iswcr.2017.01.003 International Soil and Water Conservation Research (xxxx) xxxx–xxxx T Vulević, N Dragović management domain could represent a different risk level (high, average, low), where selected management units (sub-watersheds or parcels) may be grouped The special interest of researchers attracted PROMETHEE method, designed “to be as easily understood as possible” in the 1980s (Brans & Vincke, 1985) for the decision maker to decide which action is better compared to others regarding the number of criteria The method is applied in many fields for partial or complete ranking of the alternatives, categorized in nine areas in Behzadian, Kazemzadeh, Albadvi, & Aghdasi, (2010) where are separately presented – environment management and hydrology, as well as water management This paper demonstrates the application of the PROMETHEE II method for sub-watersheds ranking in Topciderska river watershed according to soil erosion vulnerability The criteria used in the analysis are the key parameters of USLE/RUSLE methodology: topography, rainfall erosivity, land cover and soil erodibility The additional information that PROMETHEE method requires – importance of considered criteria (weights) is determined using the AHP method Sensitivity analysis is carried out to check how changing the criteria weights influence on the results Erosion potential is estimated on a cell-by-cell basis using Geographical Information system (GIS), indicating the zone with a high risk of soil erosion The results of the PROMETHEE II method and the ArcGIS application represent the valuable information in the planning stage of watershed management, where suitable soil erosion and torrent control measures should be determined and implemented Fig Topography map for sub-watersheds Model (DEM) with a grid cell resolution of 25 m Arc Hydro is used to delineate sub-watersheds in the study area through steps within the Terrain Processing toolset Firstly, stream network is generated using a threshold of 10 km2 and then sub-watersheds are delineated Material and methods 2.1 Study area 2.2 Soil erosion dataset estimation Topciderska river watershed, located in the northern part of Serbia in Belgrade (between 44°34′N-44°48′ latitude and 20°25′E-20°34′E longitude) covers the area around 147 km2 (Fig 1) Sub-watersheds boundaries are generated using Digital Elevation GIS-based soil erosion modeling is performed using topography, soil erodibility, rainfall erosivity and land cover data For all criteria, the map is produced and soil erosion risk is estimated by map overlays Topography factor (LS factor), known in USLE methodology as slope length and steepness factor, has the largest influence on the soil loss in Europe (Panagos, Borrelli, & Meusburger, 2015) LS factor is obtained by the equation of Desmet and Govers (1996) using a 25 m DEM, where the upslope contributing an area of each grid cell as well as grid cell slope have to be defined (Fig 2) Rainfall erosivity (R factor) reflects the effect of rainfall intensity on soil erosion Mean annual rainfall dataset for rain-gauge stations was collected from the Republic Hydrometeorological Service of Serbia R factor (Fig 3) is computed as mean annual rainfall (mm) multiplied by the coefficient 1.3 (Grimm, Jones, Rusco, & Montanarella, 2003; Van der Knijff, Jones, & Montanarella, 2002) The land cover (C factor) is obtained using CORINE (“coordination of information on the environment”) land cover methodology and ArcMap, where 14 Corine land cover (CLC) classes are registered (Fig 4) For every CLC class, C factor values (adopted from Panagos et al., 2015; Belanovic et al., 2011; Diodato, Fagnano, & Alberico, 2011) are assigned to all classes (Fig 5) Soil erodibility (K factor) is calculated for four soil types registered in the study area, presented on the map provided by the Serbian Institute of Soil Science (Fig 6) The estimation of K factor is performed for the topsoil and includes soil texture (content of silt, clay, and sand) and organic carbon content data, used in the EPIC model (Wischmeier & Smith, 1978; Williams, Renard, & Dyke, 1983) 2.3 PROMETHEE method PROMETHEE method (Preference Ranking Organization METHod for Enrichment Evaluations) developed by Brans in 1982 is outranking MCDA method used when someone has to decide which action is better compared to others regarding the number of criteria The method is Fig Location of the study area International Soil and Water Conservation Research (xxxx) xxxx–xxxx T Vulević, N Dragović Fig Rainfall erosivity map for sub-watersheds Fig Map with distribution of C factor value applied in many fields for partial ranking (PROMETHEE I) or complete ranking of the alternatives (PROMETHEE II), due to qualities such as a clear method with simple conception and possibilities to check the stability of the results (Brans, Vincke, & Mareschal, 1986) The complete ranking of action could be performed using the PROMETHEE II method, where necessary inputs placed into an evaluation table are: finite set of n actions A={A1,A2,…, An}, finite set of n criterion C={C1,C2,…,Cn}, which are maximized or minimized, and evaluation of actions on all considered criteria Application of the method involves five steps (Keyser & Peeters, 1996; Behzadian et al., 2010): Technique for Order Performance by Similarity to Ideal Solution – TOPSIS, Entropy), subjective approach (Simple Multi Attribute Rating Technique – SMART, Delphi method) or combinations of these two (Zadari, Ahmed, Shirazi, & Yusop, 2015) In the subjective weighting method where the AHP method belongs, the importance is given to criteria depends on the preferences of the decision maker AHP generates weights through pairwise comparisons described in Saaty and Vargas (2012) If the weight is higher (closer to 1), the criterion is more important Decrease in value of one criteria increase the value of other criteria such that ∑wj=1 Determination of preference function The preference function is representing a function of the difference between the two evaluations For each criteria, there is a preference function: Determination of the weights of decision criteria wj using available methods The relative importance (e.g weights) of criteria could be determined using: objective approach (Least Mean Square – LMS, Pj (a, b ) = Fj[dj (a, b )], 0≤Pj (a, b )≤1, (1) where, Pj(a,b) denotes preference of action an over action b, Fj is a nondecreasing function of the observed deviations dj (differences in Fig Land cover classes for sub-watersheds Fig Soil types registered in sub-watersheds International Soil and Water Conservation Research (xxxx) xxxx–xxxx T Vulević, N Dragović evaluation) between action a and b The larger the deviations, larger the preferences, which vary from to (0 for no preference or indifference, for strict preference) The preference function could be of different types: Usual, U-shape, V-shape, Level, Linear, and Gaussian Calculation of global preference index The global or overall preference index π(a, b)represents the intensity of preference of a over b, and it is calculated as n π(a, b) = ∑ Pj(a, b)wj (2) j=1 where, Pj(a,b) is preference function and wj represent weights associated with every criteria, determined in the previous steps Calculation of outranking flows Positive outranking flow ρ+(a) and negative outranking flow ρ-(a) is estimated using equations ρ+(a) = n−1 x∈A n−1 x∈A ∑ π(a, x) (3) and ρ−(a) = ∑ π(x, a) (4) where, π(a, b) is preference function calculated in step The higher the positive outranking flow and lower the negative outranking flow, the better the action is considered to be Calculation of the net outranking flow Ranking of actions is possible by determining the net outranking flow φ(a) ϕ(a ) = ϕ+(a) − ϕ−(a) + Fig Map of soil erodibility distribution value ranges from 847.61 to 921.57 MJ mm ha−1 h−1 y−1 The spatial distribution of R factor, shown on the map in Fig is produced using an IDW interpolation technique available in Geostatistical Analysis Tool in ArcMap It demonstrates that the precipitation increase with increasing altitude, so the highest value is registered in the subwatersheds 6, 7, and (Table 1, column 4) The land cover type map is obtained using CORINE methodology, where 14 classes of land cover are identified, where artificial surfaces cover 28.37%, agriculture areas occupied 46.77%, forest cover 24.73% of the study area, and the rest of the study area presents water bodies (Fig 4) The mean value of C factor value is going from (for water bodies, industrial and commercial units, etc.) to 0.30 (for non-irrigated arable land) The mean value of C factor for every sub-watershed is given in Table (column three) and spatial presented in Fig Values of soil erodibility factor that varies from 0.021 to 0.045 t h MJ −1 mm−1 (Fig 7) are assigned to soil types registered in the study area (chernozem, fluvisol, eutric cambisol, colluvial deposit and luvisol) The value of K factor is estimated for every sub-watershed, where the highest value is registered in sub-watershed 8, due to the high presence of the most erodible soil – luvisol (Table 1, column 2) The mean value of all factors is obtained using the statistics option in ArcMap and this data are placed in the evaluation table, and used as an input for implementation of the PROMETHEE II method (Table 1) using Visual Promethee 1.4 Academic Edition software Application of PROMETHEE II method requires: determining the type of criterion (maximum or minimum), weights of criterion, to choose preference function and define thresholds (preference and/or indifference) Firstly, using the AHP method, weights indicating the importance of the criterion are expressed on the ratio scale using numbers from the interval [0,1] Each considered criterion is maximized, due to the fact that the objective is to rank sub-watersheds and identify the most erodible In the second step from the six preference functions (Brans et al., 1986), we opted for criterion with linear preference (type III) for all considered factors This preference function serves to transform deviations into preference score between and 1, using indifference and preference threshold (defined by soil erosion expert) given in Table When the difference between evaluations is negligible the score is 0, otherwise, the intensity of preference increases linearly until this deviation reaches the value of preference threshold, when it is strict (Brans & (5) - where, φ (a) and φ (a) are flows obtained in step Results and discussion Soil erosion risk is usually predicted using some or all of the following factors: rainfall erosivity, soil erodibility, land cover and topography (see Ochoa et al., 2016; Welde, 2016; Vulević et al., 2015; Zhang et al., 2010; Vrieling, Sterk, & Vigiak, 2006) These factors are an integral part of the USLE/RUSLE methodology widely applied in European countries for soil erosion estimation (Martín-Fernández & Martínez-Núnez, 2011) These factors are used in this study for subwatersheds prioritization using ArcGIS and PROMETHEE II method Map layers are generated for all considered factors Topography factor in the entire watershed has a value of 0.03–14.43 (Fig 2), while the mean value for sub-watersheds is between 1.55 (registered in subwatershed 3) and 2.67 (registered for sub-watershed 7) (Table 1, column 5) This indicates that the LS factor in sub-watershed has the greatest influence on soil erosion Rainfall erosivity factor is estimated using average annual precipitation that varies from 644.71 to 725.89 mm The estimated R-factor Table Evaluations of the sub-watersheds according to K, C, R and LS criterion Sub-watershed К factor [10−2] C factor [10−2] R factor LS factor Indifference threshold Preference threshold 2.65 2.85 3.20 3.49 3.17 3.28 3.04 3.84 3.45 0.15 0.45 0.22 0.43 3.27 8.94 10.12 6.35 4.76 11.78 10.68 1.45 4.34 868.49 888.97 900.52 902.77 904.09 914.25 918.36 910.10 914.63 6.23 18.70 2.03 2.20 1.55 1.75 1.85 2.32 2.61 1.92 2.08 0.13 0.40 International Soil and Water Conservation Research (xxxx) xxxx–xxxx T Vulević, N Dragović Fig Comparison of sub-watershed ranking Vincke, 1985) For each criterion and all pairs of actions (subwatershed) preference function is estimated [Eq (1)] and used to calculate the global preference index given in Eq (2) In the final step using positive and negative outranking flow (given in Eqs and 4), the net flow is calculated and used to rank sub-watersheds (Fig 8) Sensitivity analysis based on changing weights of criteria is performed considering scenarios: 1) equal weights of all factors, 2) week preference given to K factor, 3) week preference given to the C factor, 4) week preference given to R factor, and 5) week preference given to the LS factor Week preference means weights of 0.4 for factor mentioned in the scenarios while other factors have weights of 0.20 Sub-watersheds are ranked using the net outranking flow that varies between −1 and 1, which is presented in the vertical green/red line in Fig Of the highest priority are sub-watershed with the maximal value of net outranking and positive outranking flow The ranking in case 1: > > > > > > > > 1, in case 2: > > > > > > > > 1, in case 3: > > > > > > > in case 4: > > > > > > > > and in case 5: > > > > > > > > The ranking is similar for all scenarios except for the last one, where the preference is given to the topography factor In this case, subwatershed is the highest priority, due to the high value of slope length and steepness factor and the huge amount of precipitation in this part of the watershed Sensitivity analysis shows that the most erodible sub-watershed, according to all scenarios, except in the last one (case 5) is subwatershed This sub-watershed is located in the upper part of the Topciderska river watershed, where the average amount of annual precipitation is 700.08 mm and the mean value of R-factor is estimated to be 910.10 MJ mm ha−1 h−1 y−1 Around 60% of this sub-watershed is occupied by land cover where soil erosion process could progress: non-irrigated arable land (20%), complex cultivation patterns (24.31%), land principally occupied by agriculture with significant areas of natural vegetation (18.75%) The value of LS factor varies from 0.03 to 13.74, with the mean value of 1.92 Multiplying these factors the erosion vulnerability (EV) index is obtained and used to identify the critical areas under sub-watershed These zones are presented on the map and classify into four categories: low (53.77%), moderate (29.26%), high (13.03%) and severely vulnerable areas (3.94%) (Fig 9) Around 15% of the area of sub-watershed present high and extreme vulnerable zone, which should have a priority for implementation of soil erosion and torrent control measures In Vulević et al (2015) for the same study area, 20 sub-watersheds Fig Erosion vulnerability (EI) index for the first order sub-watershed are identified and rank according to soil erosion vulnerability using three factors: mean watershed slope, soil erodibility and land cover factor MCDA methods which are applied are AHP and TOPSIS methods and the results indicate the strong correlation between ranking obtained using these methods In this study smaller number of sub-watersheds are delineated using ArcGIS, and instead of three factors, four factors are considered with a newly digitized map of soil type and land cover distribution The two sub-watersheds of the highest priority et Vulević et al (2015) are actually one part of the subwatershed 8, which outranks all the other sub-watershed applying PROMETHEE II method In addition to soil erosion factors that appear in USLE/RUSLE, some other factors can be included in the analysis, such as runoff coefficient (Krois & Schulte, 2014), sediment yield index (Kostadinov, Dragović, Zlatić, & Todosijević, 2008) etc Regarding all of this, it should take into account that quantity and quality of inputs, as well as applied methodology (criteria for watershed delineation, applied MCDA or some other methods), can influence the results Conclusions Based on the results, we can conclude that the PROMETHEE II outranking method provides a complete ranking of sub-watersheds endangered by erosion process, and thus can help decision maker to decide where to implement soil erosion and torrent control measures The results of the application depend on the available information and choose criterion, their thresholds, as well as on the preference which decision maker give to this criterion Sensitivity analysis is an available option in the Visual Promethee software which is userfriendly and available free of charge In this paper, we have opted for four factors used in USLE/RUSLE: topography, rainfall erosivity, soil erodibility and land cover, with the aim to indicate the most vulnerable sub-watershed Sensitivity analysis International Soil and Water Conservation Research (xxxx) xxxx–xxxx T Vulević, N Dragović methods European Journal of Operational Research, 89, 457–461 Kostadinov, S., Dragović, N., Zlatić, M & Todosijević, M 2008 Erosion control works and the intensity of soil erosion in the upper part of the river Toplica drainage basin XXIVth Conference of the Danubian Countries IOP Conference Series: Earth and Environmental Science, Available online at: (〈http://iopscience.iop.org/17551315/4/1/012040〉) Krois, J., & Schulte, A (2014) GIS-based multi-criteria evaluation to identify potential site for soil and water conservation techniques in the Ronquillo watershed, northern Peru Applied Geography, 51, 131–142 Macary, F., Ombredane, D., & Uny, D (2010) A multicriteria spatial analysis of erosion risk into small watersheds in the low Normandy bocage (France) by ELECTRE III method coupled with a GIS International Journal of Multicriteria Decision Making, 1(1), 25–48 Martín-Fernández, L., & Martínez-Núnez, M (2011) An empirical approach to estimate soil erosion risk in Spain Science of the Total Environment, 409, 3114–3121 Merritt, W S., Letcher, R A., & Jakeman, A J (2003) A review of erosion and sediment transport models Environmental Modelling & Software, 18(8–9), 761–799 Ochoa, P A., Fries, A., Mejía, D., Burneo, J I., Ruíz-Sinoga, J D., & Cerdà, A (2016) Effect of climate, land cover and topography on soil erosion risk in a semiarid basin of the Andes Catena, 140, 31–42 Panagos, P., Borrelli, P., & Meusburger, K (2015) A new European slope length and steepness factor (LS – factor) for modeling soil erosion by water Geosciences, 5(2), 117–126 Saaty, T L., & Vargas, L V (2012) Models, Methods, Concepts & Applications of the Analytic Hierarchy Process New York: Springer Van der Knijff, J.M., Jones, R.J.A & Montanarella, L 2002: Soil Erosion Risk Assessment in Italy In: J.L Rubio, R.P.C Morgan, S Asins & V Andreu (eds) Proceedings of the third International Congress Man and Soil at the Third Millennium Geoforma Ediciones, Logrono Vrieling, A., Sterk, G., & Vigiak, O (2006) Spatial evaluation of soil erosion risk in the West Usambara Mountains, Tanzania Land Degradation and Development, 17(3), 301–319 Vulević, T., Dragović, N., Kostadinov, S., Simić, S B., & Milovanović, I (2015) Prioritization of soil erosion vulnerable areas using multi-criteria analysis methods Polish Journal of Environmental Studies, 24(1), 317–323 Wang, X., Zhao, X., Zhang, Z., Yi, L., Zuo, L., Wen, Q Liu, B (2016) Assessment of soil erosion change and its relationships with land use/cover change in China from the end of the 1980s to 2010 Catena, 137, 256–268 Welde, K (2016) Identification and prioritization of sub-watersheds for land and water management in Tekeze dam watershed, Northen Ethiopia International Soil and Water Conservation Research, 4, 30–38 Williams, J R., Renard, K G., & Dyke, P T (1983) EPIC: A new method for assessing erosion's effect on soil productivity Journal of Soil and Water Conservation, 38(5), 381–383 Wischmeier, W H., & Smith, D D (1978) Predicting Rainfall Erosion Losses- A Guide For Conservation Planning (Agricultural Handbook537) Washington DC: USDA Zadari, N H., Ahmed, K., Shirazi, S M., & Yusop, Z B (2015) Weighting Methods and their Effects on Multi-Criteria Decision Making Model Outcomes in Water Resources Management Malaysia: Springer Zhang, X., Wu, B., Ling, F., Zeng, Y., Yan, N., & Yuan, C (2010) Identification of priority areas for controlling soil erosion Catena, 83(1), 76–86 is carried out to determine how variations of criterion weights influence on final ranking Application of ArcGIS enables spatial visualization of all factors, their intensity, and overall influence, so it is recommended to use both ArcGIS tools and PROMETHEE II method, to improve the decision-making process Acknowledgements This paper is a result of Project No 43007 (subproject No 16), funded by Ministry of Education, Science and Technological Development of the Republic of Serbia References Behzadian, M., Kazemzadeh, R B., Albadvi, A., & Aghdasi, M (2010) PROMETHEE: A comprehensive literature review on methodologies and applications European Journal of Operational Research, 200, 198–215 Belanović, S., Perović, V., Vidojevic, D., Kostadinov, S., Knežević, M., Kadović, R., & Košanin, O (2011) Assessment of soil erosion intensity in Kolubara district, Serbia Fresenius Environmental Bulletin, 22(5a), 1556–1563 Belton, V., & Stewart, T (2002) Multiple Criteria Decision Analysis: An integrated Approach Boston: Kluwer Academic Publisher Brans, J P., & Vincke, P H (1985) A preference ranking organization method (The PROMETHEE method for multiple criteria decision-making) Management Science, 31(6), 647–656 Brans, J P., Vincke, P., & Mareschal, B (1986) How to select and how to rank projects: The Promethee method European Journal of Operational Research, 24(2), 228–238 Desmet, P J J., & Govers, G (1996) A GIS procedure for automatically calculating the USLE LS factor on topographically complex landscape units Journal of Soil and Water Conservation, 51(5), 427–433 Diodato, N., Fagnano, M., & Alberico, I (2011) Geospatial and visual modeling for exploring sediment source areas across the Sele river landscape, Italy Italian Journal of Agronomy, 6, 85–92 Fernández, C., & Vega, J A (2016) Evaluation of USLE and PESERA models for predicting soil erosion losses in the first year after wildfire in NW Spain Geoderma, 273, 64–72 Grau, J B., Antón, J M., Tarquis, A M., Colombo, F., de los Rios, L., & Cisneros, J M (2010) An application of mathematical models to select the optimal alternative for an integral part to desertification and erosion control (Chaco area – Salta Province – Argentina) Biogeosciences, 7, 3421–3433 Grimm, M., Jones, E.R., Rusco, E & Montanarella, L (2003) Soil Erosion Risk in Italy: A revised USLE approach European Soil Bureau Research Report No 11, EUR 20677 EN (2002), Office for Official Publication of the European Communities, Luxemburg Ishizaha, A., & Nemery, P (2013) Multi-criteria Decision Analysis: Methods and Software United Kingdom: Wiley Sons Keyser, W D., & Peeters, P (1996) A note on the use of PROMETHEE multicriteria ... demonstrates the application of the PROMETHEE II method for sub- watersheds ranking in Topciderska river watershed according to soil erosion vulnerability The criteria used in the analysis are the key... distribution The two sub- watersheds of the highest priority et Vulević et al (2015) are actually one part of the subwatershed 8, which outranks all the other sub- watershed applying PROMETHEE II method. .. Geographical Information system (GIS), indicating the zone with a high risk of soil erosion The results of the PROMETHEE II method and the ArcGIS application represent the valuable information in the planning