AbstractInvestigation of temporal variations of groundwater resources quality and quantity is very important for the comprehensive management of these resources. In this study, trends of the groundwater quality and quantity parameters variations of the Sari-neka plain during the period (1991-2015) were investigated. For this purpose, Mann-Kendall and Sen’s slop estimator non-parametric tests at 95% and 99% confidence levels to evaluate the trends were analyzed. The parameters used in this study were Ca2+, Mg2+, K+, TDS, Na+, Cl-, So42- in the qualitative part and the groundwater level in the quantitative part. The results for 16 quantity observation wells indicated that the trend of groundwater level for 5 observation wells was positive. 2 observation wells were negative trend and 9 observation wells were a non-significant trend. The qualitative trend analysis for 19 observation wells showed that expect Ca2+ which experienced positive trends (3 observation wells), the other quality parameters followed negative or non- significant trend.

In groundwater resources management, continuous monitoring of wells discharge and charge and groundwater quantitative and qualitative variations are very important. Groundwater monitoring network could provide information on quantitative and qualitative variations in groundwater situations. Hence it would be used to predict the behavior of the aquifer in the future. This study describes an optimizing method to design the optimum groundwater-level monitoring network in aquifer of Eshtehard. Two scenarios were considered to optimize the locations of the observation wells: (1) exploring optimal locations of existing observation wells and (2) adding optimal observation wells to existing well network. Kriging method was utilized to estimate groundwater-level for preparing the database. Optimization of groundwater monitoring network with two objectives (1) minimizing root mean square error and (2) minimizing number of wells, was implemented by non-dominated sorting genetic algorithm (NSGA-II). The results show that by using this method it is possible to remove redundant wells in a monitoring network or to add new wells in case of shortage of observation wells in a network having acceptable accuracy.

Considering the complexity of groundwater media and high cost of groundwater monitoring, new innovative methods help us in the improvement of groundwater systems. In this research for minimizing overall data loss in total of optimized network, particle swarm optimization algorithm was used. In order to verify the effectiveness of the algorithm, one monitoring network with 57 observation wells was chosen. Later by reducing the number of monitoring wells to 42, the amount of RMSE was measured. Considering 0.3 as the threshold of error, the optimized network was obtained using the remaining 45 observation wells. A comparison of groundwater contours with original network showed that the solution of proposed algorithm is effective. In the remaining 42 observation wells there were some minor discrepancies. In comparison with genetic algorithm, the solution of PSO algorithm is better and has higher efficiency. Comparison of their RMSE also shows a faster convergence for PSO algorithm. Prediction of groundwater level contours in both models were possible and comparable too. Distribution of the omitted observation wells supported this comparison

Groundwater level monitoring networks already exist in most alluvial aquifers in Iran. A question needs to be answered as to what extent networks are able to provide proper information about the groundwater resources to meet the water resources management objectives. Therefore, it is necessary to evaluate the efficiency of the existing monitoring network and redesign it if necessary.  This study has attempted to assess the condition of existing observation wells and the relationship between the wells and the aquifer by combining hydrogeological information and the results of a well-video inspection. Hydrogeological studies can reveal that the observation wells were drilled in which hydro-geological sequence (hydrofacies) and the external factors influence water level fluctuations. Moreover, observation well's video inspection operations, performed for the first time in Iran, can show the condition of the well casing and screen and the water-column length. The proposed framework was applied to evaluate the groundwater-level observation  well's efficiency in Shirvan alluvial aquifer in North Khorasan province, Iran. The results showed that out of 17 observation wells in the aquifer, one well is not placed into the main aquifer, six observation wells need to be rehabilitated due to screen clogging and three observation wells are at risk of going dry.

By studying changes in the water levels of 359observation wells before and after the construction of 251 water reservoirs in Gorganroud watershed of Golestan province, water reservoirs effects were evaluated in improving the groundwater level. In this study some water reservoirs were selected that there were observation wells around them. All wells in which the water table levels were lower than the water levels of water reservoirs were evaluated. Among these, 31 Water reservoirs clearly showed there effects on 33 observation wells. Therefore, 12.4 percent of Water reservoirs have a positive effect on the level of groundwater. With providing a distribution map of all water reservoirs and observation wells and adapt them to each other and having observational data before and after the water reservoirs construction the effect of water reservoirs on the aquifer improvement were evaluated. In this study, parameters for evaluating effect of water reservoirs on the improving of aquifer included time of influencing the water level on observation wells levels, annual water rise and annual compensation of the water level. In this study the results show that water reservoirs had a positive effect on the groundwater level at the distance of 3.8 km, increasing the maximum groundwater level by 114 cm. The amount is of the annual compensation of the water level was 184 cm. Based on a rating scale the evaluation criteria were graded and so the performance of the water reservoirs at levels low (31.2%), average(20.4%), good(17.2%), good (21.5%) and excellent (9.7%) was evaluated.

Groundwater level prediction is an essential priority for planning and managing groundwater resources. This study aimed to compare the accuracy of the Neuro-Fuzzy Adaptive Inference System (ANFIS) model with the ANFIS model combined with particle swarm optimization algorithm (ANFIS+PSO) in predicting the monthly groundwater level of Golpayegan aquifer during 2002-2019. For this purpose, monthly data on rainfall, temperature, pan evaporation in the selected meteorological stations, discharge volume of exploitation wells and groundwater level of observation wells have been used. After spatial and temporal analysis, four observation wells with two input data structures (S1 and S2) were selected to predict the groundwater level. The results of trend and homogeneity tests show a 99% significance of groundwater level changes in the selected observation wells 4, 8, 19 and 20 with a sudden drop of 22, 17, 27 and 2 meters before and after June, September, July and August 2010, respectively. The highest and the lowest accuracy of groundwater level prediction is related to observation wells 20 and 4 with root mean square error values (RMSE) of 2.37 and 0.21 m, respectively, related to ANFIS_S1 and ANFIS + PSO_S2 models. Generally, the results of this study indicate that the selection of appropriate structure of input data is more effective than the combination of two models (ANFIS and PSO) in increasing the accuracy of groundwater level prediction. So, that the optimal structure of input data and the combination of optimized algorithm model have increased the accuracy of groundwater level prediction, 44% and 25%,  respectively.

Groundwater monitoring networks provide important data which are necessary to understand the dynamics of hydrogeological systems. Since the cost of the installation and maintenance of groundwater monitoring networks is extremely high, optimal design and the assessment of the effectiveness of the monitoring networks is necessary. This paper presents the application of a newly developed geostatistical method based on the acceptance probability concept, to optimize the existing network of observation wells in the Shirvan alluvial aquifer, located in North Khorasan province. To this aim, by choosing a suitable semi-variogram and using ordinary kriging, the acceptance probability in the aquifer was calculated. Then, based on the spatial pattern of groundwater level, the acceptance probability was calculated for various parts of the aquifer and the acceptance accuracy values were analyzed at different levels of probability. The results showed that based on the existing observation wells network, 19.2% of the aquifer area has a very high acceptance accuracy. On the other hand, by modifying the network of existing observation wells and adding suggested points, 46.7% of the aquifer area will have a very high acceptance accuracy. Therefore, the installation of new observation wells in the proper locations will increase the accuracy acceptance and improve the efficiency of the observation well network.

The present study presents a methodology for the design of long-term groundwater head monitoring networks to reduce spatial redundancy in which the additional wells if not sampled، the error related to groundwater level estimation would be negligible. This method is based on Support Vector Machine، and founded upon the statistical learning theory. Throughout the study، some 63 quantitative data، observation wells as well as meteorological parameters (precipitation and evaporation) of Ramhormoz plain (in a 7-year period) were employed to evaluate the performance of Least Squares Support Vector Machine model (LS-SVM) in the groundwater observation well network design concept. Different combinations of parameters affecting the ground water level were assessed using the model LS-SVM. The optimal combination of LSSVM model with RBF Kernel function carries such performance parameters as R2=0. 9992، MAE=0. 3405. Then، using Function Approximation Optimum، a number of 42 observation wells were pinpointed to apply the appropriate spatial monitoring in the plain of RAMHORMOZ.

Selection of optimum interpolation technique to estimate water quality parameters in unmeasured points plays an important role in managing the quality and quantity of water resources. The aim of this study is to evaluate the accuracy of interpolation methods using GIS and artificial neural network (ANNs) model. To this end, a series of qualitative parameters of samples from water taken from Dehgolan aquifer located in Kurdistan, Iran including CL, EC and PH were evaluated by any of the models. In this study, qualitative data from 56 observation wells with good dispersion in the whole plain was used. The data of 46 observation wells were used for calibration and the data of other 10 wells were used for verification of models. The results showed ANNs, IDW, and Kriging excellence and accuracy over other models in estimation of quality parameters CL, PH and EC. However the ANNs model is more accurate than other models. In case of lack of time and the need for acceptable accuracy and less risk in the estimation of qualitative parameters, the use of ANNs model is superior to other statistical models used.

Since the withdrawal of the observation wells at the plains done for the point, it is necessary to calculate the average groundwater level and also generalization the estimated water level from collected point to the surface of plain. The aim of this study is an investigation on the application of combined geostatistics method with optimized artificial neural networks by genetic algorithm in interpolation of groundwater level over Dezful and Zeidoon plains located in the Khozestan province. The obtained results from Cokriging, Kriging and IDW methods indicated that Cokriging with the Gaussian variogram and cross-variogram in Dezful Plain, and Kriging with the Gaussian variogram in Zeidoon plain are the best geostatistical methods for estimation the groundwater level and combined with artificial neural networks. Also the results of combination these two models showed that optimized model by genetic algorithm have better evaluation criteria than geostatistical methods and proposed as an effective combined model for estimation of the groundwater level. So that an application of this optimized combined method in Zeidoon plain with fewer observation wells was better than Dezful plain