Groundwater quality monitoring and simulation is of great importance in arid and semi-arid regions due to water scarcity in such regions. Geostatistical methods are appropriate techniques for interpolating and spatial mapping of groundwater qualitative parameters. In the present study, geostatistic-based Kriging and Co-Kriging methods were used and compared with data-driven artificial neural network (ANN) and ANFIS models for predicting spatial distribution of groundwater electrical conductivity (EC), then the best model was selected for further sampling in the studied region. Data from 24 wells in the Keshit, Bam Normashir, and Rhmtabad plains (Kerman province, Iran) during the period of 2002–2011 were used for analysis. Root mean square error (RMSE), R², and mean absolute error (MAE) statistical indices were used for assessing the applied models. Results showed that for both the models, i.e. triple input model (including longitude, latitude, and the number of months) and the quadruple-input model (including longitude, latitude, number of months, and Cl), ANN model presented the most accurate results with the lowest RMSE and MAE and the highest R² values. Attending to the geostatistical methods, Co-Kriging method outperformed the Kriging method. Finally, using the tested geostatistical model, the EC data were produced for the regions without observational values.

Wind speed is an essential component that needs to be determined accurately, especially over long‐term periods for various engineering and scientific purposes including renewable energy productions, structural building sustainability and others. In this study, six different heuristic methods: multi‐layer perceptron artificial neural networks, (ANN), adaptive neuro‐fuzzy inference system (ANFIS) with grid partition (GP), ANFIS with subtractive clustering (SC), generalized regression neural networks (GRNN), gene expression programming (GEP) and multivariate adaptive regression spline (MARS) are developed to model monthly wind speeds using meteorological input information. The atmospheric pressure, temperature, relative humidity and rainfall values are obtained from Jolfa and Tabriz meteorological stations, Iran, and are used to build the proposed predictive models.. Different statistical indicators are computed to evaluate and comprehensively assess the performance of the six heuristic methods. Over the testing phase, the ANFIS‐GP and GRNN models are seen to exhibit the highest predictive performance for the Jolfa and Tabriz stations, respectively. That is, the maximum coefficient of determination are found to be 0.874, 0.858, 0.850, 0.849, 0.847 and 0.826, for the GRNN, ANFIS‐GP, ANFIS‐SC, ANN, GEP and MARS models, respectively, for Jolfa station, respectively, revealing the superiority of GRNN over the five counterpart models. The results show the generalization capability of the tested heuristic artificial intelligence techniques for both study stations, and therefore could be explored for windspeed prediction and various decisions made in regards to climate change studies.

Soil is one of the main elements of natural resources. Accurate estimation of soil erosion is very important in optimum soil resources development and management. Analyzing soil erosion by water on cultivated lands is an important task due to the numerous problems caused by erosion. In this study, the performance of three different data driven approaches, e.g. multilayer perceptron artificial neural network (ANN), grid partitioning (GP), and subtractive neuro-fuzzy (NF) models were evaluated for estimating soil erosion. Land use, slope, soil and upland erosion amount were used as input parameters of the applied models and the erosion values obtained by MPSIAC method were considered as the benchmark for evaluating the ANN and NF models. The applied models were assessed using the coefficient of determination (R²), the root mean square error (RMSE), the BIAS, and the variance accounted for (VAF) indices. The results showed that the subtractive NF model presented the most accurate results with the minimum RMSE value (3.775) and GP, NF and ANN models were ranked successively.