M. Sakizadeh; M. T. Sattari; H. Ghorbani
Abstract
The soil samples were collected from 170 sampling stations in an arid area in Shahrood and Damghan, characterized by prevalence of mining activity. The levels of Co, Pb, Ni, Cs, Cu, Mn, Sr, V, Zn, Cr, and Tl were recorded in each sampling location. A new method known as min/max autocorrelation factor ...
Read More
The soil samples were collected from 170 sampling stations in an arid area in Shahrood and Damghan, characterized by prevalence of mining activity. The levels of Co, Pb, Ni, Cs, Cu, Mn, Sr, V, Zn, Cr, and Tl were recorded in each sampling location. A new method known as min/max autocorrelation factor (MAF) was applied for the first time in the environmental research works to de-correlate these elements before their geo-statistical simulation. The high cross-correlation among some elements, while poor spatial correlation among the others, could have made spectral decomposition of MAFs unstable, resulting in some negative eigenvalues, so it was decided to reduce the dimensionality of the original variables by Principal Component Analysis (PCA). The resultant 6 heavy metals (Cr, Mn, Cu, V, Ni, and Co) were converted to their respective MAFs followed by their geo-statistical simulation using Sequential Gaussian Simulation (SGS) independently. Examination of the cross-variograms of MAFs indicated that the resultant factors had been rigorously de-correlated, especially at zero lag and around ∆ lag distance. Several validation checks including reproduction of variograms in data and normal score space, close matching between distribution of MAFs versus simulated realizations, and reproduction of descriptive statistics and data histograms all confirmed that the data values had been honored by this applied method. The results obtained indicated that this method could reproduce the data values as well as the spatial continuity of heavy metals (e.g. semi-variograms) successfully. In addition, this technique is simpler and more computationally efficient than its equivalent sequential Gaussian co-simulation as fitting a linear model of co-regionalization (LMC) is not required in the data-driven MAF method.
M. Sakizadeh; R. Mirzaei
Abstract
The aim of this work is to examine the feasibilities of the support vector machines (SVMs) and K-nearest neighbor (K-NN) classifier methods for the classification of an aquifer in the Khuzestan Province, Iran. For this purpose, 17 groundwater quality variables including EC, TDS, turbidity, pH, total ...
Read More
The aim of this work is to examine the feasibilities of the support vector machines (SVMs) and K-nearest neighbor (K-NN) classifier methods for the classification of an aquifer in the Khuzestan Province, Iran. For this purpose, 17 groundwater quality variables including EC, TDS, turbidity, pH, total hardness, Ca, Mg, total alkalinity, sulfate, nitrate, nitrite, fluoride, phosphate, Fe, Mn, Cu, and Cr(VI) from 41 wells and springs were used during an eight-year time period (2006 to 2013). The cluster analysis was used, leading to a dendrogram that differentiated two distinct groups. The factor analysis extracted eight factors accumulatively, accounting for 90.97% of the total variance. Thus the variations in 17 variables could be covered by just eight factors. K-NN and SVMs were applied for the classification of the aquifer under study. The results of SVMs indicated that the best performed model was related to an exponent of degree one with an accuracy of 94% for the test data set, in which the sensitivity and specificity were 1.00 and 0.87, respectively. In addition, there was no significant difference among the results of different kernels, indicating that an acceptable result can be achieved by selecting the optimum parameters for a kernel. The results of K-NN showed roughly a lower efficiency compared with those of SVMs, where the sensitivity and specificity was reduced to 0.90 and 0.88, respectively, although the accuracy of the model was 93%. A sensitivity analysis was performed on the groundwater quality variables, suggesting that calcium next to nitrate were the most influential parameters in the classification of this aquifer.
