ardabil plain

The environmental consequences of land subsidence are destructive, costly and irreparable, and include creating cracks on the surface of the earth, damaging human structures such as building foundations, streets, bridges, roads, and power transmission lines. Sewage is destruction of irrigation systems and fertile agricultural soils and damage to ancient sites. Remote sensing methods, unlike mapping data and topographical maps, which are in physical contact with terrestrial phenomena, are without the slightest interference on terrestrial phenomena, and measuring and evaluating changes in phenomena are evaluated from a distance. Short receiving time and high spatial accuracy of radar images have made it used as a general and widely used tool to estimate land subsidence. According to the statistics announced in the country of Iran, the adverse effects caused by land subsidence are not a low number and are rapidly developing and spreading in different regions of the country. Leave irreparable damage. Ardabil plain, with its rich underground water resources and good soil, has always been of interest in the last half century and has been a suitable place for providing drinking water and agriculture. With the boom of agriculture from the 1960s onwards and as a result the excessive harvest from the aforementioned table since 1363, the aforementioned source began to decline and the continuation of this situation in the following years caused this plain to become more critical.

In order to carry out this research, the data of 22 piezometric wells in the Ardabil plain have been used. The time period used in this research is a 7-year period from 1395 to 1402. The method used for the data of this section is the BRF method, as one of the methods of radial functions, which is used due to its low error value and high accuracy. SAR images of the European Space Agency's Sentinel 1 satellite in SLC format and with vv polarization have been used to find out the changes in land subsidence in the Ardabil plain. The images used by the Sentinel 1 satellite (in c-band with a wavelength of 5.6 cm) are in the group of sensors with medium resolution in terms of spatial resolution. The radar interferometry method provides the possibility of producing a digital model of the ground height, whose optimal height accuracy for c-band data with a wavelength of 5.6 cm is about 5 meters. This method is able to reveal surface changes in the ground in different intervals with millimeter accuracy by using at least 2 or more radar images. In this method, an artificial interferogram is produced with the help of digital elevation model of the earth and conversion of height into phase, and in this way, with the help of reverse DEM data, the phase effect caused by topography is calculated and removed from the phase difference values. The remaining phase difference belongs to the effect of surface displacement and atmosphere.

The results of the investigation of piezometric wells in the area of Ardabil Plain show that the maximum drop in the underground water level is 48.77 meters in the southeast of Ardabil Plain and the lowest drop in the underground water level is 1.57 meters in the north. Eastern Ardabil plain was calculated. The amount of fluctuations in the water level of piezometric wells shows that the highest amount of fluctuations was in the area of Pirqavam wells, Arallovi Bozor and Khalil Abad lands. The lowest fluctuation was also observed in the area of Agchechai wells, Nojedeh. In the studied time period, the water level of Khalilabad piezometric well in 2015 was 2.96 meters, while in 1402, the water level in this area reached 46.3 meters. During 7 years, the water level has dropped by 43 meters, which indicates a critical situation in this sector. The lowest fluctuation of the water level is also for the Aghche-chai well. The land subsidence map of Ardabil plain shows that: the south-eastern parts of Ardabil city and also to some extent in the southern part have suffered ground subsidence due to the extraction of underground water. In the next order, the western parts of Ardabil city are prone to land subsidence. Based on this map, the maximum amount of ground subsidence has been calculated at around 598 mm in the area of Khalil Abad well. In the area of Khalil Abad well, the situation of underground water level drop is very critical and it has dropped by 43 meters during 7 years from 1395 to 1402. The overlapping results of the underground water level and co-depth curves with the results of radar interferometry show the accuracy of the findings in this section. Overlapping the location of the historical sites with the land subsidence map shows that, first of all, Tapraqlu hill (first millennium BC) is in the area of land subsidence with a rate of 250 mm.

The results of the application of this method revealed a very high level of land subsidence for the Ardabil plain (598 mm during a 7-year period). The southeastern areas of Ardabil plain have found a critical situation in recent years due to unprincipled exploitation and lack of proper management. Overlapping the location of the historical sites with the land subsidence map showed that Taparaglu hill with a rate of 250 mm and Ozrik Tepe-si with a rate of 69 mm are in the area of land subsidence, respectively. The three historic sites of Qala-Bovini, Niar Tepesi and Tezre Tepesi are also located in the area prone to land subsidence with rates of 27 mm, 46 mm and 49 mm, respectively. The cause of land subsidence in the Ardabil plain, according to the studies conducted on the changes in the underground water level, is the excessive exploitation of the underground water resources for the cultivation of agricultural products and providing the possibility of compaction of the underlying layers. In general, it can be said that the research results indicate the involvement of historical sites in the area of land subsidence.

Monitoring the changes and fluctuations of precipitation in geographical areas can give a better view of the behavior of this phenomenon in the coming years. The purpose of this research is to investigate the precipitation situation in Ardabil Plain (Ardabil, Bileh-Daragh, and Kolour stations) and forecast it in the coming years based on the output of CMIP6 models by the CMhyd downscaling model. Then, using R2, MAE, MSE, RMSE, and Taylor diagram, the observational data of the base period were compared with the historical data of 5 GCM models from CMIP6, and the best model was selected for each studied station. The output of the top models was corrected for skewness by linear scaling method and based on SSP126, SSP245, and SSP585 scenarios, the precipitation of 2050-2023 for each station, forecast, and its trend were drawn with the Mann-Kendall statistic. The results showed that in the eastern and western areas of Ardabil Plain (leading to the heights of Talesh and Sablan mountains), the rainfall changes were increasing (2.80 mm). In the Ardabil station, the MIROC6 model with a correlation coefficient of 0.94%, and in Bileh-Daragh and Kolour stations, the MPI-ESM1-2-HR model with a correlation coefficient of 0.88% and 0.92%, respectively, have the highest accuracy in simulating the precipitation. Also, the results of the scenarios showed that the precipitation changes in Ardabil station in the future period compared to the base period under the SSP126, SSP245, and SSP585 scenarios will be 0.24, -6.36, and -2%, respectively.

Systems for assessing groundwater vulnerability are designed to protect groundwater resources from pollution. The DRASTIC method is a well-known approach for determining groundwater susceptibility. One drawback of the DRASTIC method is that it relies on expert judgment to rank parameters, which introduces uncertainty. This study used a new generation of Fuzzy Logic (FL), called the Z-number theory, to estimate the specific vulnerability of aquifers and address this uncertainty. The specific vulnerability of the Ardabil and Qorveh-Dehgolan aquifers was estimated using two scenarios: the DRASTIC parameters as inputs and nitrate concentration values as output. The vulnerability of the aquifer was also evaluated by comparing the results of the proposed models with those of the DRASTIC model, which served as a benchmark. The analysis showed that the Z-number Based Modeling (ZBM), which considered data reliability and weighted the rules appropriately, produced higher-quality results than the classic FL. In the Ardabil plain, the ZBM yielded results that were 53% better (using seven inputs) and 184% better (using four inputs) compared to the classic FL. In the Qorveh-Dehgolan Plain (QDP), the ZBM produced results that were 127% better (using seven inputs) and 311% better (using four inputs) than the classic FL. The irregularity and non-linearity of the data, such as the high coefficient of variation (CV) in the Ardabil plain compared to the QDP, may contribute to the high CV value in the plains. Therefore, in plains with high CV, the quality of the extracted Z-number-based rules may be lower.

Therefore, the purpose of this research is to investigate and analyze the most important factors involved in creating the risk of subsidence in the Ardabil plain and to identify the susceptible surfaces that are likely to be involved in subsidence in the near future. The purpose of this research in the first stage is to evaluate the subsidence using radar interferometry technique in the Sarscape software environment, using the capabilities of A1 Sentinel images in the time frame of 2016 and 2021, and also in the following, in relation to the zoning of susceptible areas with the algorithm Aras multi-criteria was implemented in Edrisi software environment. The results of the present study showed that between 0 and 22 mm of subsidence has occurred in the studied area, and the highest amount of subsidence is concentrated in the central part and then in the eastern and north-eastern parts. According to the results of subsidence risk zoning; The criteria of water level drop, distance from the river, geology, and land use are the most important factors involved in creating the risk of subsidence in the study area, respectively, with a weighting factor of 0.221, 0.166, 0.152, and 0.147, respectively, and 267/41 and 403/21 square kilometers of the range have a very high probability of danger. Finally, it can be said that the most important factor involved in increasing the amount and potential of subsidence in the Ardabil plain is the excessive use of underground water and the drop in the water level.

One of the dangers facing Iran’s plains is subsidence. Land subsidence due to various reasons, such as over-harvesting of underground water sources and climate changes, has caused many problems in agricultural lands, roads, and power and energy transmission lines. Therefore, it is important to deal with the causes and influencing factors to control the risks. Ardabil plain is also one of the areas prone to subsidence due to climate changes and the decrease of underground water in recent years. The aim of the present research is the zoning subsidence risk in this plain. First, the factors influencing subsidence in the Ardabil plain (slope, land use, lithology, distance from the fault, distance from the waterway, drop in water level, distance from the city and the village) were identified. Then the layers of data were applied in the geographic information system. In the next stage, the weighting of the investigated factors was done using ANP method and in the Super Decision software, and the final analysis and modeling were done using the MABAC method as one of the multi-criteria decision making methods. Finally, the resulting map was classified into five categories from very low risk to very high risk. The results showed that such factors as water level drop, distance from the river and lithology have the highest weighting factor. Moreover, 244.29 and 370.59 square kilometers, respectively, of the area of this plain, are in very dangerous and dangerous classes. Finally, due to the high potential of Ardabil plain in terms of subsidence, protective and management measures should be taken into consideration by relevant authorities and institutions.

Estimation of subsidence rate using radar interferometry technique and groundwater parameters and land use (Case study: Ardabil plain)IntroductionThe phenomenon of subsidence has affected many parts of the world, including Iran, and in recent years has been raised as one of the main issues and challenges. This phenomenon has developed in different regions under the influence of different factors. In Iranian subsidence, groundwater loss has been suggested as the main factor. Because there is a direct relationship between subsidence of areas and the rate of groundwater loss in different areas. In fact, the prevailing dry climate in most parts of Iran and the concentration of industrial, agricultural and drinking water consumption on groundwater resources, has provided a good infrastructure for the occurrence of this phenomenon. Due to the importance of the issue in recent years, various studies have been conducted in the field of subsidence and advances in the field of remote sensing have led to monitoring the phenomenon of subsidence with greater accuracy and speed than in the past. . Which has received a lot of attention in recent years Radar interference method that has high accuracy and speed in processing information and monitoring land surface changes. Therefore, in this study, this method has been used to monitor the subsidence of Ardabil plain.MethodologyIn this study, First, radar interferometry method has been used to investigate and measure the amount of subsidence. One of the powerful tools for monitoring the subsidence phenomenon is the radar interferometry method. This method is able to determine the changes in the earth's surface in that time period by comparing the phases of two radar images taken from an area at two different times. Sentinel 1 satellite images from 2016 and 2020 were used for this purpose. Considering the direct effect of groundwater decline and increasing its utilization rate on subsidence, in the present study, the status of piezometric wells in the Ardabil plain was evaluated by intriguing method (kriging). Kriging interpolation requires zero mean estimation error. Absolute estimation in interpolation is one of the main features of Kriging model. Also, This means that the estimated value of the quantity at the sampling points is equal to the measured value and the variance of the estimate is zero. Object-oriented method was used to classify the images in the land use map in this study. This is because the method uses information from other pixels (size, shape, texture, etc.) for classification in addition to spectral information. Image objects are created based on parameters such as scale, object shape, color, compression ratio that are determined during interpretation.ResultsThe study of subsidence according to Figure 3 in the last 5 years from 2016 to 2020 in the study area shows that the amount of subsidence is 4 mm per year and 22 mm during 5 years. In order to reconcile the results of radar interferometry with other supplementary data, land use map as well as piezometric wells in Ardabil plain were used. By reviewing and analyzing the land use map and subsidence of Ardabil and field studies confirm the maximum occurrence of subsidence in agricultural areas and good rangelands or 0.22 mm, medium rangelands 0.21 mm and 0.20 mm It provides unsafe and risky living conditions for the residents of Ardabil plain and indirectly has reduced the quality of environmental conditions and life of the residents of Ardabil plain. Land subsidence is a pervasive phenomenon in the world, which has had a significant quantitative and qualitative manifestation in recent years, mainly due to the improper exploitation of groundwater resources and the intensification of its surface decline. To investigate the situation and the effect of groundwater level decline and its effect on the subsidence of Ardabil plain from the information of 39 piezometric wells (obtained from Ardabil Regional Water Organization) during the years 1987 to 2020 that the results of the analysis of a decrease of 48% Shows the water level in 2020 compared to 1987. The maximum water level of piezometric wells has increased from 45 meters to more than 70 meters in 2020, which indicates the deterioration of the Ardabil plain aquifer. Which requires the identification of effective factors, mitigation measures and adaptation measures, including the restriction or prohibition of groundwater extraction in areas subject to subsidence and the implementation of strict regulations for extractors of groundwater resources in areas subject to subsidence and continuous monitoring Indicators of subsidence.ConclusionsIn this study, the extent and distribution of subsidence in Ardabil plain and the effect of groundwater level decline and land use on land subsidence were investigated by radar interferometry, interpolation and object-oriented methods. Studies show subsidence of 22 mm over a period of 5 years. The results of data analysis of 39 piezometric wells in Ardabil plain in a period of 33 years show an average groundwater drop of about 9.5 meters. Co-institutionalization of subsidence maps with land use layers also confirms the maximum occurrence of the settlement area, which has the highest rate related to good pastures and agricultural areas0 / 22. Mm, medium rangelands 0/21 mm and man-made areas0 /20 mm.

The aim of this study was to investigate the factors affecting the decrease and change of groundwater level in Ardabil plain in two periods 1995 to 2005 and 2005 to 2015. The monthly precipitation data of Ardebil, Nir, Namin, Abi baglo, Hir, Samiyan stations in the Ardabil plain during the statistical period of 1995-2015 and monthly data of the height of the station in 24 Piezometric well ring were chosen for the plain. Landslide OLI and TM satellite imagery was used to prepare land use map for the target periods in June 1993, 2005, and 2015. The results of land use changes in the years 1993, 2005, and 2015 in the Ardabil plain showed the highest watery agriculture with 48156.26, 50678.66, and 58356.68 and area water level, respectively, were with 168.75 ,88.65 and 380.95 ha, lowest level Which indicates the high level of agricultural land involvement in the decline of agricultural land in the Ardebil plain. The study of the process of Piezometric Wells showed that in the plain of Ardabil, the maximum height of the surface of the station (1437 m) is related to the southern parts of the plains around the village - Noshahr-Kargan and the minimum height (1300 m) is related to the village of Khalifaulo Sheikh. The highest level of cultivation is also focused on user plans in these areas.

Streamflow forecasting is required in many activities associated with the planning and operation of water reservoir systems. In hydrology, short and long-term streamflow forecasting are essential to optimize hydrological systems, which can lead to the expansion or reduction of future projects. In this research, using wavelet-based de-noising data and wavelet transform, a temporal pre-processing approach was applied to the monthly runoff time series in Samian station at the outlet of Ardabil plain. In the second stage, the rainfall-runoff modeling was performed by Artificial Neuron Network (ANN) for three different combinations of data. The first and second combinations of input were used from Samian station data but the third combination was used from the upstream runoff data of Samian station (Gilandeh and Kozatopraghi stations). The results showed that the serves of both wavelet-based de-noising data and wavelet transform (WT) techniques could improve the performance of the ANN rainfall-runoff modeling of the Samian station up to 4% and 18.5% in the verification phase. The third combination of data demonstrated better accuracy in comparison to the other data set.

In this study, the effect of climate change on time of wheat cultivation, duration of growth period was investigated in Ardabil region. For adaptation to the phenomenon of climate change, and was developed appropriate crop calendar for wheat cultivation in Ardebil. To achieve this approach, first the occurrence of climate change was analyzed by using statistical downscalling, and the output data of the CanESM2 model using SDSM software under the RCP 4.5 scenario, next the climate parameters of the maximum and minimum temperatures and precipitation were simulated for the future period (2040-2011). Then the time of wheat cultivation was determined according to two temperature and precipitation factors for the base and future periods. The results of calculation of length of growth were obtained using GDD index. Finally, an appropriate agronomic calendar for the future years was determined based on this indicator. The results showed that the temperature is increasing and the average annual temperature in Ardebil will increase from 9.2 to 10.2 ° C and precipitation will be reduced to 15 millimeters. Under the conditions of the future climate, length of Growth period will reduce 20 days relative to the basic period. The time of wheat cultivation will be postponed for 15 days and the suitable time of wheat cultivation will be in the second half of October.

Groundwater has played an important role in the urban and rural water supply and agriculture. In order to manage water resources, an accurate and reliable groundwater level forecasting is needed. In this research, 15 piezometers in Ardabil plain were used. SVM was applied for a prediction method in one month-step-ahead. Clustering tool and Wavelet Transform (WT) as spatial and temporal pre-processing and an artificial neural system for modeling were also used. The results showed that the values of R2 coefficients in calibration and verification of prediction were respectively 0.94 and 0.89. On the other hand, the application of the WT to groundwater level data increased the performance of the model up to 3% and 5% for calibration and verification parts. The performance of the SVM model was compared to the proposed combined WT–ANN and ANN models. The results showed that the values of R2 coefficients in calibration and verification of prediction were respectively 0.94 and 0.88. The application of the WT to groundwater level data increased the performance of the model up to 3% and 7% for calibration and verification parts. The results obtained by the SVM model showed the improved performance of modeling and its combination with WT showed the best performance in the pre-processing of the modeling. Also the results of the ANN and hybrid WT-ANN models yielded good performance. Also, the results of the hybrid WT-ANN models showed slightly better results than the ANN model in some clusters. Recently, Artificial Intelligence (AI) approach, as a new generation of robust tools, has been developed for time series forecasting purpose. As such forecasting tools, Artificial Neural Network (ANN) and Support Vector Machine (SVM) have been extensively employed at different engineering fields. Among such AI models, the capability of the commonly used ANN models to approximate nonlinear mappings between inputs and outputs makes it a useful tool for modeling hydrological phenomena. However, ANN-based modeling may include some shortcomings, such as over fitting, convergence to local minima and slow training, which make it difficult to achieve adequate efficiency when dealing with complex hydrological processes [12]. Support Vector Machine (SVM), proposed in [13], is one of the most persuasive forecasting tools as an alternative method to ANN. SVM is based on the structural risk minimization principle and Vapnik–Chervonenkis theory, and involves solving a quadratic programming problem; thus, it can theoretically get the global best consequence of the primal problem.
In recent decades, SVMs have been implemented in several hydrological fields and in groundwater levels. In this paper, the conjunction of SVM and the wavelet-based data pre-processing was examined by proposed Wavelet-SVM (WSVM) in modeling groundwater level for one month ahead. The proposed models were also compared with single SVM, ANN and Wavelet-ANN (WANN) models. The plain of Ardabil (38 – 38 N and 47 – 48 E), located in the north-west of Iran, covers an area of about 990 km2. In this plain, 15 piezometers (wells) are operated to measure the GWLs. The data sampling has been reported in one-month intervals for all of the piezometers. The plain is equipped with one runoff gauge at the outlet and 6 rain gauges within the watershed. Fig. 2 shows the position of piezometers as well as rainfall and runoff gauging stations. The monthly rainfall, runoff, and GWL data were available from 1988 to 2012 and used in this study. About 18 years of data were used for the training, and the remaining 7 years for the validation.
Support Vector Machine
SVM as a powerful methodology was used for solving problems in non-linear classification, function estimation, and density estimation. Via SVM, a non-linear function can be shown as: (1)
where f indicates the relationship between the input and output, w is the m-dimensional weight vector, φ is the mapping f unction that maps x into the m-dimensional feature vector and u is the bias term.
Artificial Neural Network (ANN)
ANN is widely applied in hydrology and water resource studies as a forecasting tool. In ANN, feed– forward back–propagation (BP) network models are common to engineers. The Feed forward neural network (FFNN) is widely applied in hydrology and water resource studies as a forecasting tool. Three-layered FFNNs, which have usually been used in forecasting hydrologic time series, provide a general framework for representing nonlinear functional mapping between a set of input and output variables.
The explicit expression for an output value of a three layered FFNN is given by (Kim and Valdes, 2003):     (2)     where i, j and k respectively denote the input layer, hidden layer and output layer neurons. wji is a weight in the hidden layer connecting the i th neuron in the input layer and the j th neuron in the hidden layer, wjo is the bias for the j th hidden neuron, fh is the activation function of the hidden neuron, wkj is a weight in the output layer connecting the j th neuron in the hidden layer and the k th neuron in the output layer, wko is the bias for the k th output neuron, fo is the activation function for the output neuron, xi is i th input variable for input layer and k and y are computed and observed output variables, respectively. NN and MN are respectively the number of the neurons in the input and hidden layers. The weights are different in the hidden and output layers, and their values can be changed during the network training process.
Wavelet transform (WT)
The WT has enlarged in occupation and popularity in recent years since its inception in the early 1980s, but the widespread usage of the Fourier transform has yet to occur (Grossman and Morlet, 1984).
In real hydrological problems, the time series are usually in the discrete format rather continues and, therefore, the discrete WT in the following form is usually used (Mallat, 1998): (3)
where m and n are integers that respectively control the wavelet dilation and translation; a0 is a specified fined dilation step greater than 1; and b0 is the location parameter and must be greater than zero. The most common and simplest choice for parameters are a0 = 2 and b0 = 1. This power-of-two logarithmic scaling of the dilation and translation is known as the dyadic grid arrangement.
Self Organizing Map (SOM)
SOM is an effective software tool for the visualization of high-dimensional data. It implements an orderly mapping of a high-dimensional distribution onto a regular low-dimensional grid. Thereby, it is able to convert complex, nonlinear statistical relationships between high-dimensional data items into simple geometric relationships on a low-dimensional display while preserving the topology structure of the data (Kohonen, 1997). The way SOMs go about reducing dimensions is by producing a map of usually 1 or 2 dimensions which plot the similarities of the data by grouping similar data items together.
The SOM is trained iteratively: initially the weights are randomly assigned. When the n-dimensional input vector x is sent through the network, the distance between the weight w neurons of SOM and the inputs is computed. The most common criterion to compute the distance is the Euclidean distance (Kohonen, 1997): (4) The results of the proposed one-step-ahead GWL modeling using pre-processed data by SVM and WT-SVM were given. The SVM-based results were also compared with those of the ANN-based model.
Results of clustering
Due to the existence of various piezometers over the Ardabil plain and the importance of managing groundwater resources, it is a necessity to unite the adequate information about GWLs in various regions of the plain and identify the dominant piezometers to predict GWL conditions of the plain in the future. In order to accomplish the spatial clustering, an SOM was utilized to identify similar and predominant piezometers. The SOM classifies the similar piezometers (with similar temporal patterns and seasonalities) into the same classes.
The clustering results of piezometers into 5 clusters are shown in Table 1. It is clear that clustering was achieved in the direction of main stream flow and probably groundwater flow regime was parallel with the surface water toward the outlet in the northwest of the plain. To evaluate the performance of the clustering results produced by SOM, the Silhouette coefficient was used as a measure of cluster validity. The Euclidean distance was then utilized to select the centroid piezometer of each cluster, which was the best representation of the GWL pattern of the cluster.
Table (1) The results of clustering     Cluster NO.
Piezometers
Silhouette Coefficient
Central Piezometer  
Cluster 1
P4, P9
0.42, 0.34
P4  
Cluster 2
P2, P12
0.46, 0.72
P12  
Cluster 3
P1, P8, P11
0.45, 0.58, 0.11
P8  
Cluster 4
P6, P7, P10, P14
0.41, 0.62, 0.40, 0.54
P7  
Cluster 5
P3, P5, P13, P15
0.65, 0.71, 0.53, 0.51
P5     The results of one-step-ahead for all 5 central piezometers of clusters are shown in Table 2. As mentioned previously, for each ANN, the dominant input variables (column 2, Table 2) were determined by linear correlation, in which Pi(t) and Ij(t) respectively indicate the GWL and rainfall time series of central piezometer i and rainfall gauge of j. Q(t) is the outflow time series from the outlet of basin. The results of one-step-ahead indicated that all of the models produced acceptable outcomes, and confirm the appropriate identification of the representative GWL patterns over the watershed. Cluster 1 did not show reliable results because the Silhouette coefficient of P4 had a lower value than 0.5, which shows that cluster 1 had a weak structure.
Piezometers in cluster 3 showed better results than cluster 1, despite the large utilization in the region which was due to being close to the outlet of the plain and accumulation of water of other regions near the outlet area. Other clusters showed superior results since they were near the supplying and recharging resources and in the highlands of plain. Therefore, the spatial clustering not only can enhance the modeling performance by grouping the similar time series within the same clusters but also it can identify the piezometers and regions with irrelevant data due to artificial and/or external impacts on the system.
Table 2 Results of ANN and SVM models for one-step-ahead predictions     Cluster NO.
Input variable
Output
variable
Model Type
R2
RMSE (Normalized)  
Calibration
Verification
Calibration
Verification  
Cluster 1
P4(t),
P4(t-1),
I4(t),
Q(t)
P4(t+1)
SVM
ANN
0.977
0.977
0.958
0.951
0.006
0.006
0.005
0.005  
Cluster 2
P12(t), P12(t-1),
Q(t)
P12(t+1)
SVM
ANN
0.944
0.935
0.86
0.869
0.041
0.044
0.035
0.034  
Cluster 3
P8(t),
P8(t-1),
I3(t-1),
Q(t-2)
P8(t+1)
SVM
ANN
0.99
0.996
0.99
0.992
0.023
0.015
0.015
0.014  
Cluster 4
P7(t),
P7(t-1),
I4(t-1),
Q(t-1)
P7(t+1)
SVM
ANN
0.819
0.832
0.667
0.677
0.038
0.037
0.023
0.022  
Cluster 5
P5(t),
P5(t-1),
Q(t-1)
P5(t+1)
SVM
ANN
0.955
0.97
0.94
0.94
0.006
0.005
0.004
0.004     Results of WANN and WSVM models
In addition to spatial patterns, some temporal features may also exist in the GWL process due to highly non-stationary fluctuations of the time series. To handle such features, wavelet-based temporal pre-processed data were entered into the ANNs or SVM in order to improve the modeling accuracy. The hybrid model, Wavelet-ANN (WANN) and Wavelet-SVM (WSVM), were simultaneously designed to catch the non-linear GWL modeling. Due to the structure of the Daubechies-4(db4) mother wavelet which is almost similar to the GWL signal, it could capture the signal’s features, especially peak values, and was selected as the mother wavelet for the decomposition of the GWL time series in this study. The decomposition of the main GWL time series at level L yields L+1 sub-signals (one approximation sub-signal, Pa(t) and L detailed sub-signals, Pdi(t) (i=1, 2, …, L)). The decomposition level 3 was considered as the optimum decomposition level. The decomposed sub-series of GWL (each resolution demonstrating a specific seasonal feature of the process) accompanied by the rainfall and runoff data of each cluster were used in the FFNN and SVM models in order to predict one-month-ahead GWL values. The results of WANN and WSVM models for one-step-ahead forecasting are presented in Table 3. The WANN and WSVM results of one-step-ahead showed that the performance of models for all clusters were accurate during both training and verification periods. According to Table 3, the results obtained by the WANN model show the improved performance of modeling in comparison to the ANN modeling. It is clear from the performance criteria that all WSVM yielded slightly better results than the WANN (except for clusters 1 and 5 in scenario 2).
Table 3 Results of WANN and WSVM models for one-step-ahead predictions     Cluster NO.
Input variable
Output
variable
Model Type  
R2
RMSE (Normalized)  
Calibration
Verification
Calibration
Verification  
Cluster 1
Pi4(t),
I4(t),
Q(t)
P4(t+1)
WSVM

WANN  
0.993
0.988
0.973
0.975
0.003
0.005
0.004
0.004  
Cluster 2
Pi12(t),
Q(t)
P12(t+1)
WSVM

WANN  
0.962
0.968
0.901
0.916
0.033
0.031
0.029
0.027  
Cluster 3
Pi8(t),
I3(t-1),
Q(t-2)
P8(t+1)
WSVM

WANN  
0.997
0.997
0.995
0.995
0.013
0.013
0.011
0.011  
Cluster 4
Pi7(t),
I4(t-1),
Q(t-1)
P7(t+1)
WSVM

WANN  
0.898
0.922
0.822
0.861
0.028
0.025
0.017
0.015  
Cluster 5
Pi5(t),
Q(t-1)
P5(t+1)
WSVM

WANN  
0.979
0.971
0.967
0.963
0.004
0.005
0.003
0.003    
In this paper, ANN based models were developed for GWL forecasting over the plain of Ardabil, in the north-west of Iran. The inputs of the AI models were monthly rainfall, runoff, and GWL at 15 piezometers over the study area. Data pre-processing via SOM and WT were shown to be useful tools in improving AI based GWL forecasting models. The proposed methodology was applied to Ardabil plain data to find one-month-ahead forecasts of GWL. As a result, the entire study area was divided into five clusters with SOM clustering scheme and then AI modeling was performed separately for each cluster. In order to improve model efficiency and consider seasonality effects, the WT which can capture the multi-scale features of a signal, was used to decompose GWL time series into different sub-signals at different levels. The sub-signals were then used as inputs of the AI models to predict GWLs. Overall, the results of this study provide promising evidence for combining spatial and temporal data pre-processing methods, and more specifically SOM and WT methods, to forecast GWL values using the AI method. One of the advantages of the proposed method is that by using a clustering method it is possible to identify piezometers and regions with good and bad data quality. In order to complete the current study, it is recommended to use the presented methodology to forecast the GWL by adding other hydrological time series and variables (e.g., temperature and/or evapotranspiration) to the input layer of the model. Moreover, due to the uncertainty of the rainfall process and the ability of the Fuzzy concept to handle uncertainties, the combination of the ANN and fuzzy inference system (FIS) models as an adaptive neural-fuzzy inference system (ANFIS) model, could provide useful results. It would also be useful to apply the proposed methodology in other heterogeneous groundwater systems in order to investigate the overall effect of the climatic conditions on the performance of the proposed model.

The study of changes in water resources in each region is essential to manage water resources and using them. In this study, the goal is to evaluate the available water resources in the plain of Ardebil in terms of surface and subsurface resources based on four criteria including natural, hydrological, agricultural and humanitarian by using fuzzy network analysis. In order to assess better fuzzy network analysis evaluation, sub-criteria of population, industry condition, rainfall situation, the status of surface water (volume taken from the river) and groundwater (wells, springs and aqueducts status), the area under cultivation and the type of products in terms of water demand, slope and elevation are used. Dependencies among sub-criteria using DEMATEL fuzzy technique and according to experts are determined. Using the fuzzy network analysis all criteria and sub- criteria are weighed, and the maps for all sub-criteria, are generated in according to the weight obtained. Finally, the result map that is based on initial layers and weighted based on the fuzzy techniques is generated in GIS. The resulting map is identified the sensitivity of the study area in terms of potential water resources. The study area (Ardebil plain) is located in the northwest of Iran and is delimited by latitudes 38°05′ N and 38° 30′N and longitudes 48°15′ E and 48° 35′E. The average height is about 1360 meters from the sea level. It covers an area of about 820 km2 and is part of Qara Soo river basin. The low risk areas 11.13 % equivalent to 9200 hectares are located on the northern and a bit in west of the plain. The average risk areas 19.36 % equivalent to 15870 hectares are located in the north and west of plain. The high risk areas 21.3 % equivalent to 17510 hectares are located mostly in the central and upper parts of the plain. The vulnerable risk areas 31.9 % equivalent to 26220 hectares are located in the southern and central parts of the plain and finally the critical areas 16.1 % equivalent to 13250 hectares are scattered mostly in the south and east of the study area.

Thunderstorms pose a significant threat to modern societies and their assets. Despite their local-scale characteristics، severe thunderstorms and associated extreme events like heavy rainfall، hail، gusts، or tornadoes، cause considerable damage to agriculture، buildings، or infrastructure facilities. Thunderstorms are highly localized and largely stationary weather systems affecting a limited area of about 20–50 km2، depending on the size of the cumulus tower. They are associated with shower clouds in which electrical discharges can be seen as lightning and heard as thunder on the ground، and they represent an advanced stage in the development of convection in moist air. The importance of rainfall generated by thunderstorms lies in the fact that it is largely torrential and of high intensity، and as a result much is lost as runoff which causes flooding. Basically thunderstorms occur more frequently above land areas in the warm season، while they are more frequent in the cold season over oceans. A lot of factors impact on their occurrence. Among them the most important are the thermodynamic and kinematic states of the atmosphere، topography، land cover، and its coastal configuration and atmospheric circulation issues. Ardabil is located in the northwest of the country، for this reason it has always been under the influence of the thunderstorms. Due to the geographic location and specific local conditions in this region، every year numerous thunderstorms happen in this area and cause severe damages in the agriculture، utilities and in restructure sectors. From this point of view، studying this phenomenon in detail and identifying the synoptic patterns of the ground surface and upper levels in which lead to the occurrence of thunderstorms in Ardebil are vital and important for the region. For this study، firstly the related data of Ardabil thunderstorms had been received from the Meteorological Agency of Ardabil. Within the related codes with the thunderstorms، codes from 90 to 99 during the period of 20 years (from 1992 to 2012) were used. After the initial Ardabil thunderstorm’s data analyzing، the 88 observational days that thunderstorm occurred in it were identified، and out of the 88 days، 43 days، that were compatible with the observational hours of4 for the NOAA data (3:30، 9:30، 15، 30، 21، 30) were used، for the patterning، so that the done research not to have any time contradiction with the upper atmosphere data and to be justified with the interpretation and analysis. Then، for the patterning and extraction of patterns in the upper atmosphere and ground surface، the related data to the pressure of the ground surface and geo potential height were obtained from the site which belongs to the National Center for Environmental Prediction (NCEP). So as to do this research the environmental to circulation method was used. In this case، first based on the recorded data in the Ardabil station، the occurred thunderstorms were identified and then by using the clustering، extraction and identification of the patterns for the ground surface and upper atmosphere launched، that leading to the occurrence of this phenomenon. For the classification and extraction of the ground surface pressure patterns and geo-potential height with the level of 500 hpa، the diverse kinds of hierarchical clustering methods tested، and finally based on the results، the clustering method into the Euclidean distance was known as the best method and the results of that were reflected in the following research. Result and Discuss: According to the results of hierarchical clustering، 4 patterns on the ground surface and in the level of 500 hpa were identified that the extracted patterns could justify beautifully the Ardabil thunderstorm occurrence. The ground surface patterns that lead to the occurrence of thunderstorms include pattern 1: the formation of low pressure on India، Siberia and high pressure on northern Europe and west of China، pattern 2: the formation of low pressure on the Ganges Valley، Persian Gulf and North Europe and high pressure on Siberia and west of China، pattern 3: the formation of low pressure on India and Persian Gulf and high pressure on Siberia and west of China، pattern 4: the formation of high-pressure on the central Asia and front occurring in the northwest of Iran. Also the patterns of geo-potential height with the level of500 hpa that lead to the occurrence of thunderstorms include pattern 1: the formation of the trough on the east Mediterranean and the placement of Ardabil in east of trough، pattern 2: the formation of the omega blocking on the northern parts of the Caspian sea and placement of Ardabil in southwest of that، pattern 3: the occurring of the cut-off blocking with low pressure on the central and east of the Turkey and placement of the Ardabil in east of the trough made by that، and pattern 4: the occurring split flow blocking in central Europe and placement of the Ardabil in east of trough. Different patterns according to the results of hierarchical clustering in the ground surface and in upper atmosphere are impressive on Ardabil thunderstorm precipitation. In the ground surface، the formation of low pressure on the Ganges valley، southwest of Persian Gulf، Siberia and northern Europe، also high pressure on the northern Europe، Siberia، west of China and the Central Asia played a significant role in the thunder precipitation occurrence. On the other hand in upper atmosphere، the study area in the east of trough and the formation of different patterns of blocking (omega، Split flow and cut-off low pressure) provided the condition so the thunder precipitation in Ardabil to be occurred.

Groundwater resources are important sources for the supply of water in agriculture, industry and drinking in Ardabil plain, therefore underground water resources planning and sustainable management of these resources are important. The purpose of this study is grading the villages in the plain of Ardabil in underground water crisis and changes during the years 1360-1391. The information obtained from 39 wells, piezometers in plain of Ardabil. Using simple techniques and fuzzy cumulative weighting and interpolation methods, the piezometers interpolation of shallow water table and how it changes during the period is showd.

Groundwater is one of the main sources of drinking water supply for many people around the world, especially in rural areas. Groundwater can be contaminated by natural or human activities are numerous. All activities including residential, municipal, commercial, industrial and agriculture can affect groundwater quality. Groundwater contamination can result, such as the loss of a source of water supply, high cost of clearing the high cost of alternative water supply or cause potential health problems. Given the importance of determining the results of the plains of the country, the aim of this study was to determine changes in aquifer storage of Ardabil using statistics and analysis on multi-criteria decision-making and evaluation of groundwater is a crisis situation.
Data and

In this study, the data of piezometers wells in of Ardabil plain scattered through the city of Ardabil Regional Water Authority have been prepared. Also, the surface layers and point to the plains of Ardabil, political divisions and the location of wells, piezometers villages for final maps have been used. The data of deep wells, as well as cultivation of four major product with a high water requirement of wheat, barley, potatoes and forage to determine the relationship between ground water and water harvesting has been a drop in water table.
The study area: Plain study area is located in the north-west of Iran in Ardabil province (Figure 2 and Figure 3). The average height is about 1360 meters above sea level It covers an area of approximately 820 square kilometers and is located in the Gharasoo watershed.
Inverse Distance Weight;
Global Polynomial Interpolation;
Local Polynomial Interpolation;
Radial Basis Functions;
Straight Ranking;
Fuzzy Normalized;
Fuzzy multi-criteria decision-making;
FSAW.
The first step is to evaluate each process and required hydrological data collection, and the coordinatingits location. The geostatistical methods of IDW, GPI, LPI, and RBF in the ArcGIS software were used for interpolating all existing data and a drop in water table in the area of standards for grades 10 class (raster) within restricted fields of Ardebil were determined.
Finally, using simple collective weight, weight-bearing layers and layers of loss data water table for the years 60 and 90 is obtained. To get the final map of water table drops, the two layers are deducted and the final map of Ardabil plain water table drop that phase is obtained.
Analysis showed the reduction of water table almost 47 percent in 1391 compared to 1360. As can be seen in Figures 12 and 13, maximum of 45 meters water table wells, piezometers in 1360 to more than 70 m in 1390 has come to reveal the deterioration of the aquifer Ardabil.
Pholadloo_e_Shomali district with the highest concentration of deep wells in the near future to continue the removal of existing deep wells, groundwater resources will go into sharp decline.
Sharghi Village goes to the crisis and in the meantime, the central Vilkij district includes the eastern part of the plain, the drop in water table aquifer at greatest risk to the two villages in East and Central Vilkij.
• Due to the limitations of traditional agricultural development potential ground water;
• Increase the efficiency of irrigation, changing crop patterns of water needed to fill low-power consumption;
• Efficient use of water resources and prevent unauthorized digging deep wells to exploit the nutritional front, especially in the East and Southeast plains.

Rivers as natural channels for collecting and transport of water, has always been an important consideration for the location of human habitations. River capture is one of the most important changes in river system that may lead to changes in other parts of the system. River capturingis a geomorphological phenomenon occurring when a stream or river drainagesystem is diverted from its own bed, and flows in the bed of a neighboring stream. As hydrological systems are inherently nonlinear(Amorocho and Brandstetter, 1971), therefore using chaos theory despite its complexities seems to be one of the simplest ways to define and predict of behaviors of these systems. The concept of chaos implies a kind of regulation in the framework of an irregular trend. One of the important issues that has been emphasized in chaos theory is the occurrence of large scale phenomena as the result of small changes in dependent variables, called butterfly effect. Current study tries to evaluate the probability of occurrence of river capture in eastern part of Ardabil plain and also its future effect on the region has been discussed. This phenomenon is occurring in a basin with area of 74.31 square kilometers near eastern border of Ardabil plain.Despite the smallarea of the basin in comparison with the Ardabil plain (4621 km2), the changes in this basin could lead to changes in whole hydrological system of the plain. The study area in this research is Ardabil plain. Rivers flowing in Ardabil plain includeBalkhlichay and Ghorichayare joined in GharaSouriver. Total drainage area of the plain is 4621.48 km2.

In this study, topographic maps (1:50000), geological maps (1:100000), Digital elevation Models (DEM) and aerial photographs besides field observations were employedfor doing precise evaluations. In order to investigate and predict of river capture phenomenon, at first effective factors on erosion in twonear basins (Ardabil and near basin in eastern slopes of Talesh) were studied and then the most critical part for occurrence of river capturingwas definedusing distance measured between river branches of two basins. As geological baseof two basins are similar, therefore the rate of erosion in 2 basins must be the same, but field observations show that the erosional rate in near basin is faster than eastern part of Ardabil plain. Slope gradient could be responsible for this process. Ardabil plain has a slope of 2-28%. Eastern parts of the plainhave an average slope of 2-5%, middle parts 0-1% and in western parts up to 28% slope could be observed. In eastern part, in comparison with near basin (20%) the slope gradient is very low. Therefore slope gradient could be an important factor in capturing river in this region. In order to investigate the probability of river capturing, 35 profileswereplotted between two drainage networks branches in 2 sides of the slope, then the distance between them were measured usingArcGIS. The results show that in profiles no. 1-12 in eastern part of the basin, the low sloperesulted in high probability for river capturing. Profile 21 has the lowest distance (115 m) and is the most critical location for capturing in future. Profile no. 26,with distance of 185 m, is another susceptiblelocation for capturing. Headward erosion and river capturing will cause a large amount of input water in Ardabil plain exit from the basin. With headward erosion and capturing, rivers in Ardabil plain will find their ways to Armenia or Caspian Sea and both of these probabilities are potential threats for Ardabil plain. Also this phenomenon may have several impacts on agricultural activities and economy of the region and transportation network (Astara-Ardabil road) and especially Ardabil airport that is located near the critical point. Furthermore, humanactivities that increase the landslide may facilitate the headward erosion and could increase river capturing process.

Evaluation of erosional chaos in Ardabil plain implies that headward erosion in eastern part of the plain will have several impacts on the regionand with increasingheadward erosion, the occurrence of the capturing in eastern slope will be inevitable and hydrological system and drainage network will be changed totally and many constructions and roads and also Ardabil airport, will be in danger. To avoid raising this problem in future, more considerationsand planning need to be established and controlling activities in critical points is necessary.

The Ardabil city area has more than 2398 km2 with 504946 residuals. The procedure of increasing of people and ratio of urban land use according to the drought (especially 1371 to 1387), immigration of rural people to the Ardabil and so becoming Ardabil as a province center in 1372 are intensified. Water consumes per capita in the developing and semi arid zone estimated between 150 to 200 liter. Water consume per capita in the Ardabil city in per day 10098920 liter/day and annually of them about 398610580 liter(368610/58 m3) are estimated. In addition water necessary of Ardabil plain in the Agriculture, industrial section’s 588000000 m are estimated. The total water potential (grand and surface waters or runoffs) is estimated about 256000000 m3. To day the water deficient of plain is 347000000 m3 is the more important factors in the subsidence of Ardabil plain. Extensive extraction of water, Climatic change, ratio of people increasing, industrial developing and tectonic variation respectively have vital roll in the subsidence of the Ardabil plain.