mohammad isazadeh

In arid and semi-arid regions, meeting the environmental need of rivers and natural lakes is drastically diminished due to the climate changes, drought, and mismanagement of the water supply and distribution in agricultural activities. The closed basins, like Lake Urmia, are more sensitive to reducing the environmental need, climate change and drought. In Lake Urmia basin with arid and semi-arid climate, water shortage and high demand for water consumption in agricultural activities have caused to increase the free accessibility of the extraction of the surface water and groundwater resources. Therefore, in arid and semi-arid regions, it is necessary to implement an integrated dynamic system (IDS) for water consumption management in all sectors of water consumption, including drinking, industrial, and agriculture needs to stabilize the groundwater resources; hence, the optimal use of water resources to increase production capacity per unit of water consumption is a fundamental solution to meet the nutritional needs of humans and improve the economic conditions of operators. The multiplication of greenhouse agricultural products compared to the open-field cultivation for the same water consumption is one of the crucial solutions in the optimal use of water resources.

The Ajichay basin with an area of 12600 km2 is one of the greatest sub-basins of Lake Urmia. The average annual temperature of this basin is 11.3oC and the average annual precipitation is 320 mm. The soil and water assessment tool (SWAT) model was used to evaluate the impacts of the implementation of development scenarios of greenhouse towns. The Soil and Water Assessment Tool (SWAT) is considered as a comprehensive hydrological model to evaluate the impacts of different factors such as climate changes, land use variations, changes in irrigation methods, agricultural management, and greenhouse development on runoff volume and groundwater level fluctuations. In this study, the SWAT model has been calibrated and verified using a 30 m digital elevation model (DEM), a soil map with 217 homogeneous zones, 5 land use maps from 1987 to 2016, data gathered from 10 hydrometric stations and more than 50 meteorological stations, hydrological studies and field measurements. In the present study, two scenarios were adopted according to the current policies of greenhouse towns development in Iran and an ideal scenario (third scenario) was also considered for the development of greenhouse towns. The first scenario for the development of greenhouse settlements was adjusted using new water resources with the aim of increasing the production of agricultural products in the region. The second scenario is based on the removal of part of the arable lands of each aquifer and the use of harvested water from the removed lands in the development of greenhouse settlements. The third scenario was based on the average production statistics in open and closed environment in Ajichay catchment, in order to maintain the current production per hectare. In modelling the Ajichay basin, the time periods of 1987-1991, 1992-2009, and 2010-2016 have been assigned to the model adaptation to the basin conditions, calibration, and verification, respectively. The statistical criteria comprising correlation coefficient (CC), Nash-Sutcliffe efficiency (NSE), and root mean square error (RMSE) have been utilized to evaluate model performance for stream flow simulation in hydrometric stations.

Statistical criteria indicated the high accuracy of simulation for all ten hydrometric stations. At Akhula station (watershed outlet), the statistics including correlation coefficient (R), Nash-Sutcliffe efficiency (NSE), and root mean square error (RMSE) were 0.92, 0.83 and 6.48 m3 s-1, respectively, during calibration, and 0.86, 0.73 and 3.23 m3 s-1, respectively, during verification periods. Development of 1875 hectares of greenhouse towns in the Ajichay watershed caused an average depletion of 11.68 m and 4.41 m within the first and second scenarios, respectively; in contrast, for the third scenario, an increase of 3.86 m for the average groundwater level of the aquifers of the Ajichay watershed was acquired rather than the initial condition. At the end of the simulation period of the third scenario, the flow volume increase of 28.5 MCM y-1 was observed at the Akhula hydrometric station. The results revealed that the development of greenhouse towns using new water resources can increase agricultural crop production and also intensify the downward trend in groundwater levels. The second scenario has caused a negative balance in the aquifers due to the eliminating the water withdrawal under the traditional irrigation system from the hydrological cycle. Only in the third scenario, the removal of nearly ten hectares of farmlands for the development of one hectare of greenhouse has caused the aquifers of Ajichay basin to have an increasing trend of groundwater level. The results of the third scenario revealed that 24% of the total reduced volume of groundwater abstraction, was discharged into Lake Urmia through rivers, and 38% was related to the water withdrawal of the traditional irrigation, and the other part was spent on increasing the water abstraction from groundwater and rivers for the agricultural purposes.

The results of the implementation of the scenarios to develop greenhouse towns revealed that the type of policies and how to implement the development of greenhouse towns can enhance agricultural production capacity as well as decrease/increase the flow of rivers in the Ajichay basin. The adverse effect on the groundwater resources occurs when the administrative institutions' approach is focused only on the increasing production capacity. Furthermore, when the ratio of replacing traditional farmlands with greenhouse towns is not commensurate with the corresponding irrigation efficiencies, it applies more stress to the groundwater resources to meet the evapotranspiration need of the greenhouse plants. The positive effect on aquifers and increased river flow occurs when the actual evapotranspiration of the plants cultivated in greenhouse towns is less than the eliminated traditional farmland.

Greenhouse crops production is one of the major strategies for managing optimum use of water resources due to creating desirable conditions for plant growth over the year, multiplying production and reducing water usage. In the present study, two scenarios were adopted according to the current policies for the development of greenhouse towns in Iran and an ideal scenario (third scenario) was also considered for the development of greenhouse towns. The soil and water assessment tool (SWAT) model was used to evaluate the impacts of the implementation of development scenarios of greenhouse towns. Statistical indicators showed very high accuracy in simulating hydrometric stations of study area, as for Akhola hydrometric station (basin outlet), the statistics of correlation factor, Nash-Sutcliffe efficiency (NSE), and root mean squared error (RMSE) in the calibration period were 0.77, 0.62, and 6.21 m3 s-1, respectively, and in the validation period were 0.83, 0.66, and 3.09 m3 s-1, respectively. The development of greenhouse towns with an area of ​​1875 ha in Ajichay basin for the first and second scenarios resulted in an average drop of 11.68 m and 4.41 m in the groundwater level of the aquifers in Ajichay basin compared to the initial conditions. Simulation of the third scenario increases the groundwater level of aquifers of Tabriz, Azarshahr, Damaneh Shomali Sahand, Bostanabad, Duzduzan, Mehraban, Bilverdi, Asbforoshan and Sarab by 4.12, 2.73, 1.45, 8.88, 10.93, 2.90, 4.79, 2.99, and 3.31 m, respectively, and also has compensated a large amount of their negative balance. The results revealed that the development of greenhouse towns using new water resources can increase agricultural crop production and also intensify the downward trend in groundwater levels.

Lake Urmia, as one of the most important lakes in the country, has an inappropriate environmental condition due to excessive consumption of water and climate change. The study of climate change and rainfall distribution in this area can improve water management in this basin. In this study, the information of 65 weather stations in the period of 1997-2016 were used for precipitation zoning of Urmia lake basin. For this purpose, the data of each month were standardized and arranged in a matrix with dimensions of (n*m) in which n is the number of stations (65) and m is the number of months (12). Principal Component Analysis (PCA) was performed on data matrix and the main components were determined according to their Eigen values greater than one. Then the principal component score (PCS) values were calculated for the selected components. These values were used as inputs in the Ward cluster analysis method. Then, the Procrustes method was used to determine the index stations. The results showed that the first two main components incorporated more than 87% of the all data variances. Based on the selected components, six distinct precipitation regions were identified throughout the basin. Moreover, it was found that four stations located in different points of the Urmia lake basin namely Mehmandar, Sarab, Babaroud and Santeh can be considered as indicator stations. These stations incorporated more than 84% of the all data variances of basin stations. The Mann-Kendall trend test showed that the rainfall in the autumn season has a significant increase trend, while annual precipitation has only a significant increase trend in one of the clusters.

Pure anatase TiO2 nanoparticles with various Ag and Sn contents were synthesized by hydrothermal and sol-gel low-temperature methods. Structural and morphological characterizations of synthesized nanoparticles were performed by X-ray diffraction (XRD), transmission electron microscopy (TEM), scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDX), and N2 adsorption/desorption isotherm and brunauer-emmett-teller (BET) techniques. The effect of synthesis procedure on the crystalline structure, crystal size, surface area, pore size distribution and photocatalytic activity of synthesized samples were studied. The photocatalytic activity was tested vs. degradation of methylene blue (MB) under black light radiation. Ag/Sn-TiO2 nanoparticles synthesized by hydrothermal method showed higher photoactivity during the degradation of MB under black light irradiation because of being enhanced in the specific surface area, total pore volume, and its reduction in the crystallite size. An artificial neural network (ANN) comprising four input variables (mol% of dopant ions, photocatalyst dosage, initial dye concentration, and pH of the solution), eight neurons and an output variable (degradation efficiency %) was optimized, tested and validated for MB degradation by Ag/Sn-TiO2 nanoparticles synthesized via hydrothermal method. The results showed that the predicted data from the designed ANN model are in good agreement with the experimental data with a correlation coefficient (R2) of 0.979. A 98.9% photodegradation efficiency of MB was achieved by utilizing 0.07 mol% Ag and 0.03 mol% Sn co-doped TiO2 at pH = 12.

Having information about qualitative and quantitative parameters distribution of groundwater supplies is one of most important parameters in integrated groundwater management. Thus, in this study it has been attempted to determine a proper model and input combination for estimation of quality parameters including electrical conductivity (EC), calcium (Ca) and sodium (Na) ions in aquifers of Guilans plain. In this study, the data from 132 observation wells during 2001 to 2013 were used and artificial neural network (ANN) and support vector model (SVM) were applied. In the first approach, estimations were conducted according to five different combinations, including water level, distance from see, total precipitation of six months and coordinates of observation wells. In the second approach, estimations were conducted based on combination of the selected qualitative parameters of gamma test with combinations of the best input in the first part. Comparison of the results from the first part indicated that SVM model outperformed the ANN mode in the estimation of Ca, Na and EC parameters. Support vector machine error values for estimating Ca, Na and EC variables at the test period were 1.218 (meq/l), 0.867(meq/l), and 175.742 (µmos/cm), while for artificial neural network these values were 1.268 (meq/l), 0.933 (meq/l), and 186/448 (µmos/cm) respectively. The results from this part showed that adding the distance from see input improves the estimation of models in all cases. In the second part, using gamma test for measuring the nine quality parameters, the best combination of quality parameters was determined to estimate the three parameters: Ca, Na and EC. The results from the second part show that both ANN and SVM models have an excellent performance in the estimation of the three qualitative parameters. ANN model error values in estimating Ca, Na and EC variables in validation period were 0.662 (meq/l), 0.305(meq/l), and 47.346 (µmos/cm), while these values were 0.671 (meq/l), 0.356 (meq/l), and 55.412 (µmos/cm) for SVM model respectively.  Obviously, the results from ANN model in this section were better than those from SVM model. Results showed that both ANN and SVM models have a great ability in predicting qualitative parameters in the aquifers. Also, in less inputs, the results of SVM model are better than those of ANN model and in more inputs it is vice versa. Results of the second section showed that gamma test is fully practical and accurate in determining the effective input combinations.

Accurate estimation of river flows is one of the fundamental activities in water resources management of river basins. Artificial neural network (ANN) and support vector machine (SVM) are the most important data mining models that can be considered for this purpose. Due to the data-based attribute of these models, probability distribution of data may have a considerable effects on their performance in river flow simulation. In order to, Zarrineh Rud River basin was selected as a study area and the investigations were done for three hydrometric stations located in this basin. In this regard, first monthly runoff probability distribution of stations were studies based on Shapiro- Wilk test and then normalization of data distribution were done. Then the performance of ANN and SVM models in monthly river flow simulation of three stations was evaluated for initial observed and normal data. Based on the results of this study, the values of 0.71, 5.93 (m3/sec), 0.80, 6.58 (m3/sec) and 0.82, 22.9 (m3/sec) were obtained for correlation coefficient (CC) and root mean square errors (RMSE) indicators in the ANN model for Safakhaneh, Santeh and Polanian stations respectively in the testing period. In the SVM model, the values of 0.70, 6.34 (m3/sec), 0.78, 7.02 (m3/sec) and 0.79, 24.31 (m3/sec) were obtained for these indicators in the mentioned stations respectively. The results showed that in river flow simulation by ANN model values of CC increase 6%, 14% and 11% and RMSE values decrease 9%, 19% and 6% for Polanian, Santeh and Safakhaneh stations respectively in the testing period due to normalization of data probability distribution. For SVM model, due to normalization of data probability, CC value increases 10% and RMSE value decrease 16% only for Santeh station. Also the results showed that the ANN model with normal input data has high performance in estimation of monthly river flow compared to the SVM model in each of the three hydrometric stations.

Selecting a model that simulate the runoff with high accuracy and less error, can be helpful in favorable management of water resources plans and increasing the performance of these plans. Also, increasing the accuracy of runoff simulation in the basins with no meteorological data, is of great significance in efficient management of water resources in these basins.

Rivers flow prediction in river basins has an important role in the operation and correct management of water resources. Determining type and number of estimator models inputs is one of the important steps in rivers flow prediction. Therefore, The Procrustes analysis (PA) method for determining the number of effective inputs was used. In this study, flow prediction was done using the flow data obtained from the Safakhaneh and Santeh hydrometric stations. The Artificial Neural Network (ANN) and The Support Vector Machine (SVM) models was used for flow prediction. The best estimation of flow is done using the MLP and SVM models in Safakhaneh hydrometric station with RMSE equal to 5.68 (m3/s) and 4.85 (m3/s), respectively, and CC equal to 0.73 and 0.78, respectively. While in Santeh hydrometric station RMSE was equal to 6.44 (m3/s) and 6.36 (m3/s) respectively, and CC was equal to 0.78 and 0.79 respectively for MLP and SVM models. PA-SVM model showed better results than SVM model in estimating Safakhaneh hydrometric stations flow with RMSE equal to 5.45 (m3/s) and CC equal to 0.73 during the test period. The results also indicated that SVM and PA-SVM models estimated the flow of Santeh station with RMSE equal to 6.85 (m3/s) and 7.03 (m3/s) respectively. Basically, results indicated that the Procrustes analysis method can be used as one of the Efficient and suitable methods for determining the number of effective inputs. Comparison of the ANN and SVM results indicated that ANN model has more accuracy than SVM model.

Evaporation is a fundamental component of the hydrology cycle and has an important role in water resources management. Daily evaporation is an important variable in reservior capacity, rainfall-runoff modeling, crop management and water balance. Measurement of actual evaporation is almost impossible, but evaporation can be estimated using several methods. There are two general viewpoint to estimation of evaporation: direct and indirect methods (Tezel 2015, Terzi 2013). It is inoperative to measure the evaporation by direct methods in all locations. Direct methods usually used proximate reservoirs or irrigation projects. The indirect methods of evaporation estimation need to various input data that are not easily available. Moreover, the evaporation have very complex and nonlinear process that simulation of its complex process using simple methods is impractical. In recent years, the artificial intelligent methods such as Artificial Neural Network (ANN) and Support Vector Machine (SVM) have been successfully utilized to modeling the hydrological nonlinear process such as rainfall, precipitation, rainfall-runoff, evaporation, temperature, water quality, stream flow, water level and suspended sediment, etc (Tezel 2015). Therefore this research evaluates the performance of ANN and SVM models in daily evaporation estimation.
The daily climatic data, air temperature, wind speed, air pressure, relative humidity, rainfall, dew point temperature and sun shine hours of Tabriz and Maragheh synoptic stations are used as inputs to the ANN and SVM models to estimate the daily evaporation. For this purpose, 75 percent of the daily evaporation data were selected to calibrate the models and 25 percent of the data were used to validate the models. Different combinations of seven input and then individual inputs applied in order to evaporation estimation.
ANNs are parallel information processing systems consisting of a set of neurons arranged in layers. These neurons provide suitable conversion functions for weighted inputs. In this study, we used Multilayer feed-forward perceptron (MLP) network. The MLP trained with the use of back propagation learning algorithm. The back-propagation training algorithm is a supervised training mechanism and is normally adopted in most of the engineering applications. The neurons in the input layer have no transfer function. The logarithmic sigmoid transfer function was used in the hidden layer and linear transfer function was employed as an activation function from the hidden layer to the output layer, because the linear function is known to be robust for a continuous output variable. The optimal number of neuron in the hidden layer was identified using a trial and error procedure by varying the number of hidden neurons from 1 to 20. In recent years, SVM as one of the most important data-driven models, has been considered in this regards. This model is a useful learning system based on constrained optimization theory that uses induction of structural error minimization principle and results a general optimized answer. The SVM is a computer algorithm that learns by example to find the best function of classifier/hyperplane to separate the two classes in the input space. The SVM analyzes two kinds of data, i.e. linearly and non-linearly separable data. For a given training data with N number of samples, represented by (x_1,y_1 ),…,(x_N,y_N) , where x is an input vector and y is a corresponding output value, SVM estimator (f) on regression can be represented by: f(x)=w.∅(x)
Where w is a weight vector, b is a bias, “.” denotes the dot product and ∅ is a non-linear mapping function. Typically, three kernel functions, radial basis, polynomial and linear are applied in SVM that use of each function with various parameters for evaporation estimation may have different results. Therefore, it is necessary to evaluate the accuracy of each of these functions and select the appropriate kernel function for evaporation estimation. Two performance criteria are used in this study to assess the goodness of models fit, which are: Correlation Coefficient (CC) and Root Mean Square Error (RMSE).
In this paper, ten different combinations of seven inputs and then individual inputs applied in order to estimate the evaporation. Results of evaporation estimation in Tabriz station indicate that the first and eighth combinations have minimum RMSE and maximum CC in test period of ANN and SVM models, respectively. Also results of evaporation estimation in Maragheh station indicate that the first and Seventh combinations have minimum RMSE and maximum CC in test period of ANN and SVM models, respectively. The ANN model using first combination including air temperature, wind speed, air pressure, relative humidity, rainfall, dew point temperature and sun shine hours climate data, achieves to the amount of 2.12 (mm) and 0.78 for RMSE and R statistics in test period for Tabriz station. The SVM model using eighth combination including wind speed, air pressure, relative humidity, rainfall, dew point temperature and sun shine hours climate data, achieves to the amount of 2.17 (mm) and 0.78 for RMSE and R statistics in test period for Tabriz station. Evaporation estimation of Maragheh station using ANN and SVM models, respectively achive to 1.62 (mm) and 1.43 (mm) for RMSE statistic in test period. In next step, individual input results show that ANN model has better estimation of evaporation values in Tabriz station and SVM model in Maragheh station. Also results indicate that the SVM and ANN models have better estimation of evaporation values using individual inputs including average temperature and sun shine hours compaired to other inputs, respectively.
The results of used models indicate that both ANN and SVM models have acceptable performance in evaporation estimation. Evaluation results show that the average temperature is better input than other six parameters in estimation of evaporation. The investigations of this study indicate that although there is no significant difference in the results of three kernel functions of support vector machine, but the Radial Basis kernel function has high accuracy and better performance in estimation of daily evaporation in comparison to other kernel functions.

Accurate prediction of river flow has an important role in the optimum management of available water resources. In recent years, support vector machine (SVM) that is one of the most important data-driven models, has been considered in this regards. This model is a useful learning system based on constrained optimization theory that uses induction of structural error minimization principle and results a general optimized answer. Such as other data mining models, the SVM model can also be used for runoff simulation when the only available data is runoff (autoregressive simulation). Typically, three kernel functions, namely, radial basis, polynomial of degree d and linear are applied in SVM that use of each function with various parameters for river flow estimation may have different results. Therefore, it is necessary to evaluate the accuracy of each of these functions and select the appropriate kernel function for runoff simulation. Since time series models, namely, AR, ARMA and ARIMA are the main models for autoregressive simulation of runoff, relative accuracy of kernel functions can be investigated by comparing their performance with these models. Therefore, assessment of the accuracy of kernel functions for monthly river flow simulation and comparison of their performance with time series models is main aim of this study.
In this study Kherkherehchiy river basin was selected as the study area and observed monthly river flow of this basin in the Santeh gauging station were applied for calibration and validation of models. For this purpose, first 75 percent of monthly river flow data (1367-1384) were selected to calibrate models and 25 percent of data (1385-1390) were used to validate models. Next, probability distribution of monthly river flow data in Santeh station were studies based on Kolmogorov-Smirnov and Shapiro- Wilk test and then normalization of data distribution were done. After optimization of parameters for each kernel functions the monthly flow values were predicted in Santeh station and the performance of these functions were evaluated using root mean square errors (RMSE) and the correlation coefficient (CC).
The investigations of this study indicated that although there is no significant difference in the results of three kernel functions, but the polynomial kernel function of degree 4 with CC and RMSE values of 0.86 and 5.88 (m3/sec) respectively in the testing period, has high accuracy and better performance in prediction of monthly flow in comparison to other kernel functions. Also the results showed that ARMA(6,2) with CC and RMSE values of 0.82 and 6.47 (m3/sec) respectively in the testing period, has good performance in prediction of Kherkherehchiy monthly flow compared to the other time series models.
Finally, the predicted monthly river flow using polynomial kernel function of degree 4 (as a representative of SVM model) was compared with the results of ARMA(6,2) (as a representative of time series model) and this conclusion was obtained that the SVM model has a better performance than time series models in the monthly river flow prediction of the Kherkherehchiy basin