Introduction: The scour at the downstream of the structure may cause structural instability and finally structural damage. Therefore, it is necessary to estimate and predict the depth of scour downstream of structures before constructed. Empirical equation to predict the depth of scour is always has error and reduces the accuracy of the results. Therefore, in last decade, the method of combining models has been used to increase the accuracy of predictions in different sciences. Instead of choosing the best single model for a specific condition, which is a traditional task, it is recommended to use a single model combination method, which will result outputs of the combination model is better in all conditions. The purpose of this study is to estimate scour depth using different combination methods by combining empirical equations (single). The single equations were also compared before and after bias correction. Methodology: In this study 306 data set are used, including 264 laboratory and 42 field data. Randomly, 75% (230 data set) of the total data were choose for training and the remaining 25% (76 data) were choose for testing the combination models. Different technique including Shu and burn, EWA, GRA, BGA, AICA, BICA, KNN and LS-SVM have been used to combine single model. Bias correction has been performed to each model before using combination models. It was determined by the ordinary least squares estimator (OLS) using training data set in each model. Results and discussion: In this study bias correction was perform on single model. In general, the slope and intercept of the single equation indicate that the scour depth predicted by a single equation is greater than measured scour depth. The best estimation before bias correction is Mason and Arumugam and the National Institute of Hydraulic Laboratory Science and Technology (National Institute) equation. The National Institution's equation is chosen as the best single model before bias correction. After bias correction, the error of all single equations has been reduced and Mason and Arumugam equation with correlation coefficient of 0. 74 and error value of 0. 23 m has the highest correlation and least error among single equations. The error values of the Machado, Martins, and National Institute equations are approximately the same, with very little difference (about 0. 01) with the Mason and Arumugam equation. The results showed that after bias correction the Mason and Arumugam predicted scour depth more accurately and selected as the best single equation after bias correction. The equations of Martins, Machado, National Institute, Mason and Arumugam, D’ Agostino and Ferro were considered as inputs (independent variable) and scour depth as outputs (depended variable) of combination methods. Before bias correction, the correlation coefficient and error of the direct weighting methods showed that the GRA method has the least error in predicting scour depth (RMSE = 0. 25) and the W2 method has more error than this method (RMSE = 0. 31). Comparison of direct weighted combination methods with single equations before bias correction showed that the GRA method has much less error than the best single equation (National Institute). The AICA and BICA combination methods provided the best estimate before bias correction in indirect weighting methods and the results are similar to the best single equation before bias correction (R2 = 0. 70, RMSE = 0. 87). All three indirect weighting methods produce approximately the same results after bias correction. The error of indirect weighting methods decreased about 70% after bias correction compared to pre-correction. The results showed that artificial intelligence combination method (LS-SVM) scour depth prediction after bias correction are similar to the results before bias correction. Conclusion: Due to the scour depth uncertainty estimation by the empirical equations, the purpose of this study was to estimate scour depth downstream of structures using combination of empirical equations (combined methods). The National Institute and Mason and Arumugam equations was selected as the best single (empirical) equations before and after bias correction, respectively. The accuracy of combination methods increased because of low accuracy of single equations before bias correction, but after bias the accuracy of combination methods did not much change with single equations. Comparison of direct weighting methods showed that GRA is the best method and has much less error than the best single equation before bias correction, but after bias correction the EWA method is the best combination method and its almost similar to the best single equation after bias correction. The results of the artificial intelligence method (LS-SVM) were same as the local weighting method before and after bias correction. LS-SVM was able to greatly increase the accuracy of the estimation by combining the single equation before bias correction, but after bias correction the effect of the combination of the individual relations was reduced and the scour depth estimation as same to the single equation.

Introduction: Water conflict is a major challenge that, if left unmanaged, will become a security issue. Although tensions over water have increased, conflicts over shared water resources are more likely to happen. The study aimed to investigate water conflict and its management strategies among farmers. Methods: The descriptive-survey research method was used. The data-gathering tool was the questionnaire, which its validity was verified through face validity. The study population included farmers who used shared water wells to provide water for agriculture (N=478). Using Cochran's formula, the sample size was 214 farmers who were selected by the simple random sampling method. Data were analyzed using SPSS software. Findings: The results showed that “, drought”,and “, increasing number of farmers”, , with an average score of 3. 56 and 3. 45 respectively on a scale of 1 to 5, are considered as the main causes of agricultural water conflict. From the farmers’,view, the priority for reducing water conflicts was the participation of farmers in managing water wells and negotiating with farmers around the water. On a scale of 13 to 65 with an average of 38. 51, the perceived agricultural water conflict was at the medium level. By increasing farm distance from the well, area of agricultural rental land, and annual income from non-agricultural activities, the perception of agricultural water conflict increased. However, by increasing owned agricultural land area and agricultural income, the perception of agricultural water conflict decreased. The main strategy used by farmers to manage agricultural water conflict was “, control”, , in which coercion and force are used to manage conflict. The “, problemsolving”, and “, avoidance”,strategies were the second and third priorities respectively.

Let G = (V (G), E(G)) be a simple, finite and undirected graph of order n. A k-vertex weighting of a graph G is a mapping w: V (G) → {1, . . ., k}. A k-vertex weighting induces an edge labeling fw: E(G) → N such that fw(uv) = w(u) + w(v). Such a labeling is called an edge-coloring k-vertex weighting if fw(e) ̸ = fw(e′ ) for any two adjacent edges e and e′ . Denote by μ ′ (G) the minimum k for G to admit an edge-coloring k-vertex weighting. In this paper, we determine μ ′ (G) for some classes of graphs.