جستجوی مقالات مرتبط با کلیدواژه « ANN » در نشریات گروه « جغرافیا »

Erosion is one of the main factors restricting the soil fertility and dust production, in several parts of the world, including Iran, has effects on climate agriculture, and human health. Controlling wind erosion would be more effective once sufficient information concerning the effective factors is available. Soil Erodible Fraction (EF) is one of the soil properties that shows the sensitivity of soil particles to wind erosion. The current research aimed to utilize ANN methods and integrating it with GA in order to estimate the soil erodible fraction to wind erosion. Allahabad plain in the southwest of Abiek city in Qazvin province is considered as one of the areas sensitive to wind erosion with strong wind direction from southwest to northeast. The drying up of Allahabad wetland will intensify wind erosion in the region and turn it into a crisis. Determining the extent of land erodibility and identifying its factors affecting can be the basis of a comprehensive plan for soil protection and land sustainability and prioritizing its implementation steps. The present study was conducted to use artificial neural network methods and combine it with genetic algorithm to estimate the soil erodible factor.

In the study area, which was part of the Allahabad plain in Qazvin province, between the coordinates of 50°15 ́- 50°57 ́ east longitude and 35°53 ́- 35°57 ́ north latitude, 95 samples were taken from 10 cm of soil surface. In the samples, the percentage of aggregates with a diameter of less than 0.84 mm as an indicator of EF and percentage of clay, sand and silt, soil saturation capacity, pH, EC, SAR, equivalent calcium carbonate (CCE) and organic matter were measured as input to the models. In this paper, to model the EF using early available characteristics, two methods of artificial neural network (ANN) and its integration with genetic algorithm (GA-ANN) were employed in order to optimize the weights. In this regard, the data were primarily divided into three categories as follows: 60% of the data series was allocated to training, 20% to validation, and 20% to network testing. In this study, MLP networks were used to model the artificial neural network in estimating the values ​​of soil erodible Fracion. In this structure, each artificial neural network includes inputs and hidden and output layers. During the learning process, the degree of network learning by the objective functions was regularly evaluated and networks with the lowest error rate were accepted. To determine the optimal network with the highest level of performance of all stimulus functions defined in the software (axon hyperbolic tangent, axon sigmoid, axon linear hyperbolic tangent, axon linear sigmoid, axon bias, linear axon and axon) by trial and error The most results were used. Levenberg-Marquardt training functions were used to teach defined networks. In this study, genetic algorithm was used to find the optimal point of complex nonlinear functions in combination with artificial neural network (GA-ANN). The genetic algorithm optimizes the weights of the artificial neural network. In fact, the objective function of the genetic algorithm is a function of the statistical results of the artificial neural network.

The results showed that the erodible fraction of soil with five soil properties including pH, electrical conductivity, SAR, clay and organic matter, had a significant correlation at the level of one percent. The models used did not have an appropriate accuracy in estimating EF in both training and testing stages, so that the highest R2 was obtained in the artificial neural network model (0.49) with test series data. Both models were slightly overestimated and the GMER values ​​for the ANN and GA-ANN models were 1.15 and 1.08, respectively, but according to the AIC index, both models had similar predictive power. Sensitivity analysis of the data showed that the greatest effect on EF in the ANN model was related to organic matter (4.07) and in the GA-ANN model was related to clay (8.14).

In the current research, the relationship between soil chemical characteristics and EF might be attributed to their previous effects on vegetation in the region. Additionally, regional evidence indicates the same finding. The highest correlation was observed between EF and soil organic matter. Based on the sensitivity analysis, in the neural network model, the greatest effect on erodible fraction was related to organic matter, pH, and EC, respectively. The effect of pH and salinity on EF could be interpreted due to their effects on vegetation and consequently, the effect of vegetation on aggregates.  An important issue in the research was that the proposed models, which were ANN and its integration with GA for estimating the soil erodible fraction, were not efficient enough for obtaining the highest coefficient of determination (R2) in the model in the neural network in the test phase (R2 = 0.49), which has an accuracy of less than 50% for estimating EF.

The Earth's atmosphere (atmosphere) is divided into concentric layers with different chemical and physical properties. To study wave propagation, two layers called the troposphere and ionosphere are considered. The troposphere is the lowest part of the Earth's atmosphere and extends from the Earth's surface to about 40 kilometers above it. In this layer, wave propagation is mainly dependent on water vapor and temperature. Unlike the ionosphere, the troposphere is not a dispersive medium for GPS signals (seeber, 2003). As a result, the propagation of waves in this layer of the atmosphere does not depend on the frequency of the signals. The delay caused by the troposphere can be divided into two parts of hydrostatic delay and wet delay. The hydrostatic component of the tropospheric delay is due to the dry gases in this layer. In contrast, the wet component of tropospheric refraction is caused by water vapor (WV) in the troposphere. The study of atmospheric water vapor is important in two ways: First, short-term climate change is highly dependent on the amount of atmospheric water vapor. Water vapor has temporal and spatial variations that affect the climate of different regions. Second, long-term climate variation is reflected in the amount of water vapor. Obtaining water vapor using direct measurements and water vapor measuring devices is a difficult task. Radiosonde and radiometers are used to directly measure atmospheric water vapor, but the use of these devices will have problems and limitations, for example, the maintenance cost of these devices is expensive and also these devices do not have a suitable station cover. The best way to get information about water vapor changes indirectly is to use GPS measurements. GPS meteorological technology can provide continuous and almost instantaneous observations of the amount of water vapor around a GPS station.Estimation of precipitable water vapor (PWV) and water vapor density using voxel-based tomography method has disadvantages. The coefficient matrix of tomography method has a rank deficiency. Initial value of water vapor must be available to eliminate it. Also, the amount of WV inside each voxel is considered constant, if this parameter has many spatial and temporal variations. In this method, the number of unknowns is very high and it is computationally difficult to estimate (Haji Aghajany et al., 2020). To overcome these limitations, this paper presents the idea of using learning-based models. To do this, in this paper, 3 models of artificial neural networks (ANNs), adaptive neuro-fuzzy inference system (ANFIS) and support vector regression model (SVR) have been used.

Due to the availability of a complete set of observations of GPS stations, radiosonde and meteorological stations in the north-west of Iran, the study and evaluation of the proposed models of the paper is done in this area. Observations of 23 GPS stations were prepared in 2011 for days of year 300 to 305 by the national cartographic center (NCC) of Iran. Out of 23 stations, observations of 21 stations are used to training of models and observations of the KLBR and GGSH stations are used to test the results of the models. In the first step, the observations of 21 GPS stations that are for training are processed in Bernese GPS software (Dach et al., 2007) and the total delay of the troposphere in the zenith direction (ZTD) is calculated. It should be noted that for every 15 minutes, a value for ZTD is calculated using the observations of each station. In the second step, the zenith hydrostatic delay (ZHD) is calculated. By subtracting ZHD from ZTD, zenith wet delay (ZWD) are obtained. ZWD values are converted to PWV values. The obtained PWV values are considered as the optimal output of all three models ANN, ANFIS and SVR. Also, the input observations of all three models will be the latitude and longitude values of each GPS station, day of the year and time.

After the training and achievement of the minimum cost function value for all three models, the PWV value is estimated by the trained models and compared at the location of the radiosonde station as well as the test stations. The mean correlation coefficient for the three models ANN, ANFIS and SVR in 6 days was 0.85, 0.88 and 0.89, respectively. Also, the average RMSE of the three models in these 6 days was to 2.17, 1.90 and 1.77 mm, respectively. The results of comparing the statistical indices of correlation coefficient and RMSE of the three models at the location of the radiosonde station show that the SVR model has a higher accuracy than the other two models. The average relative error of ANN, ANFIS and SVR models in KLBR test station was 14.52%, 11.67% and 10.24%, respectively. Also, the average relative error of all three models in the GGSH test station was calculated to be 13.91%, 12.48% and 10.96%, respectively. The results obtained from the two test stations show that the relative error of the SVR model is less than the other two models in both test stations.

The results of this paper showed that learning-based models have a very high capability and accuracy in estimating temporal and spatial variations in the amount of precipitable water vapor. Also, the analyzes showed that the SVR model is more accurate than the two models ANN and ANFIS. By estimating the exact amount of PWV, the amount of surface precipitation can be predicted. The results of this paper can be used to generate an instantaneous surface precipitation warning system if the GPS station data is available online.

For considering the effect of drought stress on some morphological and biochemical changes in two pumpkin species, Cucurbita maxima L. and Cucurbita pepo L. a kind of experiment was done by field culture in three water dispersal levels with field capacity, 2/3 field capacity, and 1/3 field capacity, based on the factorial design in random block form with four replications. The results indicated that increasing the stress level, leaves water potential under drought stress decreased in comparison to the control sample in both species. But with increasing drought stress, root length increased too. Also during drought stress, root soluble carbohydrates content, ascorbic acid content, dehydroascorbic acid, catalase, polyphenol oxidase, and peroxidase enzymes activity increased significantly in 5% level, according to the results with increasing the stress, soluble carbohydrates content decreased in leaf.

As technology evolves in many countries around the world, the need for energy is increasing, which is especially important in developing countries such as Iran, because sustainable energy is needed to develop the process of sustainable development. Due to the geographical location of Iran and having a large number of sunny days, using solar power at the scale of the power plant helps provide sustainable energy. According to the radiation map provided by the Iranian New Energy Organization in East Azerbaijan province, there is enough potential to build a solar power plant. Due to the ability of neural networks to solve complex problems, the present study has used a combination of spatial decision-making system, GIS environment, and artificial neural networks to identify potential areas for solar power generation. The data used in the study include solar radiation, precipitation, sunshine hours, temperature, altitude, slope, LULC, distance from roads, and distance from cities. Based on these criteria, training data were obtained and trained using the Levenberg-Marquardt training algorithm of FFB, CFB, and MLP networks.According to the results, the CFB network, with form 9,6,1 and RMSE values 0.084 and 0.061 for training and test data, was selected as the most suitable network and with the results obtained from this network, the location was determined. The results were classified into five classes, with about 15% identified as very favorable for the construction of photovoltaic solar power plants in East Azerbaijan Province.

In this research, we are looking for zoning of photovoltaic solar power plants using an artificial neural network in East Azarbaijan province. Since artificial neural networks need training data to perform the calculations, the criteria are weighted first through ANP, and then by using the weights obtained for the criteria, the training layer for network training is created. Using the training layer, all three FFB, CFB, and MLP neural networks have been trained to obtain the appropriate network and optimal structure.Environmental criteria are selected based on the parameters of the construction of photovoltaic solar power plants. Given that the locating process is a multi-criteria decision problem between different parameters and criteria, therefore, the software must be selected that supports both the vector model and the raster model. It also can implement multi-criteria decision-making rules. Based on this, ArcGIS 10.6 software was used for data preparation, layer preparation, and integration. Super decision and Matlab software have also been applied to the process of analyzing network decision making and artificial neural networks.

The structure of neural networks is such that by changing the number of hidden layers and its neurons, the change of the stimulus function and training algorithm of the network structure is changed and affects the output of the model. Therefore, determining the optimal structure of the network is based on trial and error, and using the evaluation criteria and comparing the results, the optimal model is modeled with the least error. However, we should be careful that if the error rate is very close to zero in the evaluation of the training results, there is a possibility of over-fitting, which means that the network created will only be suitable for the training set and adding new data will not yield a satisfactory answer.Matlab software was used to simulate the structures of different artificial neural networks and determine the optimal structure. For the present study, three FFB, CFB, and MLP neural networks with different structures have been created so that all three networksemploy the Lonberg-Marquardt training algorithm with back-propagation error (trainlm). The number of neurons ranged from 1 to 15 and the number of repeats between 10 and 700. For the FFB and CFB networks, the tansig and purelin transfer functions and for the MLP network, the hardlim and hardlims transfer functions are investigated.According to the simulations, the optimal CFB network structure is 9,6,1 with 9 input neurons and 6 middle neurons, with MSE and RMSE values for the training data 0.006, 0.084 and for the test data 0.004, 0.061, the optimal FFB network structure as 9,5,1 with 9 input neurons and 5 middle neurons, with MSE and RMSE values for training data 0.11, 0.107 and for test data 0.012, 0.111 and the optimal structure of the MLP network as 9,9,1 with 9 input neurons and 9 middle neurons, with MSE and RMSE values for the training data 0.007, 0.085 and for the test data 0.006, 0.079 have been selected. Based on these results, the CFB neural network with the structure of 9,6,1 has the best performance among the networks. For this reason, the photovoltaic solar power plants in East Azarbaijan province have been located with this network.The final map was classified into five descriptive classes using the results obtained. According to the classification, about 7.7% were in the very undesirable class, 28.4% in the undesirable class, 27.6% in the middle class, 20.6% in the desirable class, and 15.5% were in the very desirable class.

With the advancement of industry and the development of new technologies, population growth in many countries of the world has increased the consumption of electricity. Also, all developed and developing countries have realized the fact that to maintain their international status, they need to provide sustainable energy, especially electricity, from non-fossil energy sources. As societies become more aware, the limitations and harms of using fossil fuels have become more apparent, forcing countries to source some of their electricity needs from other energy sources, such as renewable energy sources. Iran, like all developing countries, is no exception. Due to the geographical location of Iran and having 300 sunny days, the use of solar energy in both large and small sectors contributes to sustainable energy supply.In this study, the authors have tried to combine the existing methods for location namely the use of spatial decision-making systems and GIS, to use new methods such as artificial neural networks to identify potential areas for the construction of photovoltaic solar power plants in East Azarbaijan province. To accomplish this, based on the criteria for the construction and location of photovoltaic solar power plants, environmental factors include solar radiation, precipitation, sundial and temperature as climate criteria, elevation and slope as physical and land use criteria, distance from roads, and distance from cities are considered as economic criteria. Based on these criteria, training data was obtained through ANP, and along with this data, and LM training algorithm was performed to train FFB, CFB, and MLP networks.Based on the MSE and RMSE evaluation criteria, the CFB network with the structure of 9,6,1 was selected as the most appropriate network and the results were obtained from this network. After preparing the final map, it was determined that solar photovoltaic power plants could be built in the province.

Paveh Township has long been affected by landslides due to specific geological and geomorphologic features and anthropic activities. This study aimed to map landslide risk and its relationship with factors affecting their occurrence and compare the ANN model with (OWA) method to assess landslide risk in Paveh Township. Therefore, landslides in the area were first identified using extensive field surveys. Maps of factors affecting landslide occurrence (lithology, slope, slope direction, elevation, precipitation, land use, distance from the waterway, distance from the road, distance from the fault, soil) in GIS software then extract the relevant layers Was done. To perform the OWA model, weighting was performed by the fuzzy method using the Critical and Evaluation and Standardization of benchmark maps and to perform an artificial neural network (MATLAB) software. Each neural network parameter was determined by trial and error method. Then with the final structure of the network with 8 neurons in the input layer, 13 neurons in the hidden layer, and 1 neuron in the output layer. According to the results of the study of slope factors, land use, lithology, and soil, respectively, by weight factor; 0.156, 0.143, 0.139, and 0.131, received the most importance. Which according to the model output (OWA) was 15.53 and 26.64%, respectively, in two very high and high-risk classes, respectively. Due to the output of the neural network 19.88% and 29.82% of the area is located on the high-risk floor. Very high-risk and high-risk areas are mainly located in 15-30% slope, agricultural use, unbearable and weak quaternary structures, and in soils with a high percentage of clay, silt, and marl. The two models were compared and the OWA model had higher accuracy.

Urban development and growth has become an uncontrollable process in most countries across the world, as it can be claimed that more than half the world’s population live in cities now. Urbanization has widely influenced the environment in local, regional, and global scale. Growing urbanization has caused some problems like landfill. Moreover, many cities in Asian developing countries face serious problems in landfill management. Thus, by growing number of people in developing countries, landfill management has become one of the major issues today. Finding an appropriate site for landfill makes an important part of the planning process. The growing rate of population and development of industrial and commercial activities and services have led to the production of vast amounts of waste in cities. Golestan province was separated from Mazandaran in 1997 to form a new province with the center of Gorgan. Since then, its population started to grow and landfill turned to be a major challenge among other things.2. Literature reviewMirabadi and Husseinabadi (2017) studied Landfill Site Selection in Bukan using Analytical Hierarchy Process (AHP) and concluded that the regions between Bokan and Simineh in the southern part of Kani Shaqaq village is the best place for landfill. Ziarri et al. (2013) studied the best location of landfill using Analytical Hierarchy Process (AHP) in Jolfa city and concluded northwestern part of the city is the best place to landfill.Celiker and Yildiz (2019), evaluated the site selection of solid waste landfill using multi-criteria decision analysis and geographic information systems in the Elazığ city, Turkey. The results revealed that the landfill suitability index values for the selected site range between 2.64 and 6.10. The major part of the landfill site has relatively low index values implying that the selected site is suitable for solid waste landfill.Al-Karadaghi et al. (2019) in Sulaymaniyah, Iraq, used multicriteria decision-making methods (WLC) and GIS for landfill site selection, and seven appropriate sites for landfill were suggested. All of these sites adopted the scientific and environmental criteria.

Gorgan is a city located in northern Alborz heights which covers an area of over 10883 Hectares (Jahani Shakib et al., 2018) and it ranks 4th among the cities of Golestan province. In order to find a location for After categorizing the factors,questionnaires were classified from 1 to 9 by experts. Number 1 had the lowest score and number 9 had the highest score. Then, expert choice was used for weighting the indicators in AHP Model. At the end, in Artificial Neural Networks (ANN), each of the indicators was fuzzied first, and then the artificial neural network was implemented.

Results show that according to experts, in AHP Model, slope and geology are top priorities and distance to fault, height, and distance from airport have the lowest priority. In AHP Model, areas in the North-east and parts of Southern area as well as areas located in the middle belt of the city tend to be more appropriate for locating landfill; because  they are far from water wells, faults, villages, the city, airport and the river and the elevation of these areas are suitable for landfilling.The previous site for landfill, located around Hezar Pich in Gorgan, has not been a suitable place according to AHP Model. According to ANN Model, Northern areas of Gorgan are inappropriate for landfill because they are both close to the city, village, airport, and surface water networks and are geologically improper for having young alluvium and alluvial fans. It must be noted that Hezar Pich is not an appropriate site for landfill according to ANN Model as well.

 This study saught to locate the best site for landfill in Gorgan and had a look at previous site of the city as well. This was achieved using 11 criteria and geographical data focusing on AHP and ANN techniques.Results obtained from 2 models AHP and ANN revealed that the most inappropriate sites for landfill were Northern areas of the city due to small distance from underground water wells, airport, cities, villages, asphalt roads and unsuitable geology. While appropriate sites for landfill, according to AHP and ANN, were areas in the North-west, North-east, middle belt of the city, and some Southern parts of the city. It is noteworthy that Hezar Pich area was improper for landfill used in the past. Waste material is currently buried in Western part of the province (Aq Qala). This factory is located 40 Kms away from center of Gorgan, somewhere between two cities of Aq Qala and Gamish Tappe, covering an area of 80 Hectares. These waste materials are transferred to the factory to be recycled and processed and finally converted to organic compost.

The impact of drought on different parts is not the same. In a situation where different regions of the country have experienced a significant decline in rainfall, its impact on water resources is still unclear or the decline of surface water resources has no effect on agricultural production (Satari et al., 1395).
Increasing or decreasing in hydrological time series can be described by changes in precipitation factors, evaporation, temperature, and the like (Nourani, 1395).  Evaporation modeling from the reservoir level is important to predict the evaporation rate from the surface and the amount of water lost through evaporation and evacuated water and to have a proper planning to reduce the amount of this evaporation and its economic estimation. The heavy volume of computations and their time-consuming performance, especially in phenomena such as sudden floods, cause many financial losses and annoyances every year. One of these utilized and intelligent tools is artificial neural network which reaches acceptable output by establishing appropriate relationships between input variables in the shortest possible time and establishes the relationship with the output tool and provides the best possible result to experts. (Rajaei et al., 2010). In this regard, studies have been conducted in the world, including the study of the effect of different compounds of climatic parameters on the evaporation losses of the dam reservoir (Deswal & Pal, 2008). - Meteorological data routing nonlinear
Before proceeding to discuss the modeling and selecting the optimal model for the regions under discussion, the best nonlinear fittings are [1]obtained from the parameters affecting evaporation. For the study area, the fitting diagram for temperature data (oC), rainfall (mm), wind speed (Km/h), lake surface area (Km2) and evaporation (mm) were used, which resulted in the results and relationships for each of them.
- Introducing Artificial Neural Network
An artificial neural network consists of three main layers of the input, the hidden (middle layer) and the output layers. The layer where the results of the model analysis are generated and the modeling is done is the output layer of the model (Fig. 1). The middle layer acts as the processor of the model and the processor nodes are at this stage (Traore et al., 2010).
Fig.1 Artificial Neural Network structure with input, output and intermediate (hidden) layers - Artificial Neural Network Modeling for Minab Dam
In order to use the artificial neural network, data from the Minab Dam was estimated from the data of the years 1998 to 2014 in MATLAB software. The best structures for the neural network are given in Table 1:Table.1 Error and correlation coefficient obtained by artificial neural network
No
correlation coefficient
MSE  (Test)
MSE  (Learn)
Neural network structure
1
0.8849
0.001
0.0016
ANN(3,7,1)
2
0.8849
0.00092
0.0014
ANN(4,10,1)
3
0.89
0.00088
0.0015
ANN(4,11,1)
- Training data
For modeling of the neural network, 80% of the data was randomly selected by the MATLAB software. One of the most important diagrams used in neural network modeling is the actual values graph and evapotranspiration values using artificial neural network for training data (Fig. 2).
Fig.2 Diagram of observation data and modeling at training stage, ANN [5,5,1]
- Test data
The remaining 20% of the data was also used to test the model obtained by the artificial neural network (Fig. 3).
Fig.3 Diagram of observation data and modeling at testing stage, ANN [5,5,1] Evaporation, as one of the natural parameters, has always been of interest to experts and researchers due to the high role that human has in reaching the outflow of water. In this research, we tried to evaluate the accuracy of this model by using the artificial neural network model in estimating evaporation from the lake level of the Minab Dam. In order to investigate the evolution of the evaporation parameters for the 19-year data, the best-fit nonlinear regression was drawn and the general trend of evolution of the effective parameters was studied. For modeling of the evaporation using artificial neural network, 19-year-old statistics between the years 1995 and 2013 were used.
The best structure for estimating evaporation from the level of Minab Dam is selected in this paper. In this structure, the first and second layers have 5 neurons with 1000 replications to get the best result. The statistical coefficients obtained from the analysis using artificial neural network were considered in selecting the best structure. In this structure, the correlation coefficient with the value of 0.8941 had the highest value and the error values of training and testing the data were respectively 0.0011 and 0.0082. After this structures, ANN (3, 7,1), ANN (4,10,1), ANN (4,11,1), ANN (5,3,1) had acceptable correlation coefficient values and error in determining the amount of evaporation from the Minab dam.
[1]- MSc Student, Faculty of New Sciences and Technologies, University of Tehran.

Nowadays, investigations and analysis of changes in land use through local and national scales have been taken into consideration more than ever, the major purpose of which is to optimize land use as well as its limited finite resources.
During the past few years, the rapid growth in population along with urbanization have intensified the significance of land use, resulting in extensive changes regarding the usage of lands in the cities. In general, it has been indicated that natural forces as well as human activities are the two major factors in changing land use and ground cover through scales ranging from local to national.
Although the development of cities in western countries have taken a rather slow pace of progress currently, yet, statistics show a rapid, considerable growth in Asia. Similarly, our country have also witnessed such an increasing, accelerated progress, especially during the past 40 years; in most cases, however, the growth has taken place in outskirts of cities and locations with fewer facilities and features.
Among various changes in land use brought about by humans, urban growth and development is of utmost importance concerning the high values of lands; as a result, the position of modelling and predicting the changes in land use in the future have also gained significance for urban management, environment, and other authorities and researchers involved.
In many studies concerning the modelling of changes in land use, multi-temporal satellite images are presented as the most important type of data to be used. Consequently, the examination on how to utilize their various capabilities as to achieve desirable results is of special importance.
Tehran metropolitan area, as the political and economic capital of the country, has been subject to a rapid influx of population along with inconvenient and unbalanced development. The emergence of an entangled, polluted and overcrowded environment in Tehran have revealed the necessity of an optimized management of natural resources as well as proper use of lands in this city, more than before. It is evident that in this regard, urban designers and environment experts would draw a set of strategies in order to accomplish the aforementioned purposes.
The main goal of the present study is to investigate and analyze the effective factors in the development process of Tehran metropolitan area since 1990 until 2010, as to develop and present a model through which urban development can be predicted; such predictions would provide the basis for implementation and execution of urban planning policies. The high urban population growth rate and the lack of basic infrastructure, on the one hand, and the increasing trend of land-use disparate changes, on the other hand, clearly reveal the need for the analysis of these changes. In this regard, urban designers and environmental experts are considering strategies for the optimal management of natural resources as well as the proper use of land valuable in urban areas. The main purpose of this study is the consolidated employment of a mixed model of automatic cells and neural network algorithm based on spatial information system in order to create a model of urban development with regards to Tehran metropolitan area during time intervals of 1990, 2000, 2010, as well as comparing the accuracy of modelling in this algorithm with common models of automatic cells.
A set of parameters including the distance from urban areas, roads, streets and parks along with the slope and altitude of lands, have been considered as the effective parameters on urban development and growth. The results showed that the consolidated employment of automatic cells model and neural network algorithm can offer improvements to the calibration process of rules regarding the transfer of automatic cells.
Satellite images used in this study have been taken in 1990, 2000, and 2010. Furthermore, the required layers of information such as the maps of elevation, slopes, and land use layer of Tehran metropolitan area have been used in a shapefile format in ArcGIS 10 software. The entire processes of satellite images have been carried out in ENVI 7.4 software.
The Thematic Mapper (TM) image in 1990 and the Enhanced Thematic Mapper Plus (ETM+) images in 2000 and 2010, which would form the input and output of the models in both consecutive periods, have been classified into urban, arid, parks and farmlands, lakes and road regions using Support Vector Machine (SVM) algorithm. Then, by extracting binary image of developed regions (urban, non-urban), the spatial objects related to such regions are obtained.
In the next step, these objects are transferred to cellular space via an inverse conversion of vector space so that they could be used in the model structure while maintaining their unique identifiers. Moreover, a Cellular Automata (CA) model based on artificial neural network optimization algorithm is implemented in order to obtain the likelihood of development (transformation from non-urban to urban mode) for each cell. The model for calculating such possibility is based on mixed automatic pattern.
It is worth mentioning that the Kappa statistical index and the overall accuracy are used to assess the results and compare them to the ground truth. Furthermore, in order to validate the results, a statistical test based on variance to measure the meaningfulness of the results is utilized. Considering the limitations in common cellular patterns and vectors of automatics, the present study offers a mixed automatic pattern as a combination of cellular computing structure as well as optimal features of vector patterns. The major problem with conventional models of automatic cells entail a sensitivity to scale along with being far from the reality of ground objects. Although object-based vector models have lessened such deficiencies to some extent, yet their implementation and calculations still face a number of complexities and challenges.
In a mixed model, space is defined as a set of arranged cells, yet spatial objects derived from ground truth is also used along with them. In order to avoid employing a trial and error approach in determining the proper weights regarding the model’s components, an artificial neural network optimization method have been used to calculate the possibility of extension based on the distance from developmental factors such as the distance from roads or important central regions of the city. A statistical comparison of the ground truth of Tehran in 2010 via the simulations obtained from mixed model along with common cellular pattern shows the higher accuracy of the proposed model relative to cellular model.
The result of the study showed that the consolidated employment of automatic cell models as well as artificial neural network optimization algorithm can offer improvements to the calibration process of the rules concerning the transfer of automatic cells. A statistical comparison of the ground truth of Tehran in 2010 using simulated images taken from the mixed model and also, the common model of automatic raster cells indicate the higher accuracy of the proposed model, in a way that according to the results of modelling based on two images, the kappa index and overall accuracy for the mixed model have been estimated as 76% and 90.96%, and for the common raster model, they have been 70.47% and 87.85%, respectively. Furthermore, according to modelling based on three images, the kappa index and the overall accuracy for mixed model and also the common raster model have been estimated as 69.18%, 84.88%, 63.37% and 82.98%, respectively. The results of the study are briefly summarized in the following:Investigation of the procedure regarding the spatial-temporal changes of phenomena such as the development of cities, require employing dynamic patterns in time. In this regard, due to their simple, yet dynamic structures as well as having strong spatial features, cellular automata have been used extensively in these types of modellings.
The proposed model of mixed automata simultaneously utilizes the optimal concept and features of the vector pattern along with the simple structure and calculations of cellular pattern. Having a proper accuracy in simulations, the present model involves low sensitivity to scale and does not require complex calculations and implementation, making it an appropriate, applicable pattern in modelling of urban development.
Neural network algorithm can be used as a proper method for optimal indication of the intensity regarding the participation of various factors in regulating structures of cellular automata, rather than employing a deficient, time-consuming method such as trial and error.

In this study, Neural Fuzzy Inference System (ANFIS) was used in landslide zoning in the Songhur Chai River Basin. In order to assess the neural network, 124 occurred landslide data identified from aerial photographs, satellite imagery, and field observations and was presented to the system. In addition, for processing landslides in MATLAB software, 8-layers were prepared; slope layers, aspect, DEM, lithology, hydrographic network layer, NDVI, soil and landslide groups and landslide distribution were drawn from field studies, topographic and geologic maps and satellite images in Arc GIS software. These layers were normalized based on the largest value for each layer in the range between 1 and zero. During the modeling process, 80% of the data were selected for training and 20% for were tested and were processed in the neural fuzzy inference system. In several studies, the value is considered acceptable. Then, the values in order to map the landslide in the structure of ANFIS were processed and analyzed. Finally, with respect to the output weights, landslide zonation maps were drawn into five categories: very high, high, medium, low and very few. The results indicated that the geological structure formed of gray man and red sandstone, volcanic ash and tuff and high humidity, makes Ganjgah Mountains and Islamabad a high potential area for landslide occurrence

Linking resolution global climate models with local scale is a micro climatic process that itself is a significant issue. Recently, attempts have been made by the climatology community to develop dynamics and statistical downscaling methods for expressing climate change has taken place at a local and regional scale. Two general techniques are used for downscaling of the output of general circulation models (GCM). The prior is using of statistical methods in which the output of a statistical model (MOS) and a planned approach to weather short-term numerical prediction is presented. The second is regional climate model (RCM), that same is limited GCM model in a subnet of the network global model and by dynamic method uses climatic conditions temporal changes according to GCM model. Both methods Play an important role in Determine the potential effects of climate change caused by increased greenhouse gas emissions. Much work is done to use this method for downscaling of the global model output in different areas In which the performance of the model is assessed and uncertainty analysis has been done on these methods or were compared by other statistical methods.
In this study, three approaches to statistical downscaling methods are provided. The first approach uses random generation of climate models based on time series and Fourier series delivers. LARS-WG statistical model(Rskv et al., 1991, 27) is one of the ways is built on the basis of this approach,. In this model, the empirical distribution of daily series of dry and wet precipitation and solar radiation use is desirable. The minimum and maximum daily temperatures as the daily stochastic process with mean and standard deviations are taken daily. Seasonal cycles by means of finite Fourier series are of the order of 3 models and model residuals (model errors) is approximated by a normal distribution.
The second approach is regression model or transfer function that is more used, which uses the relationship between atmospheric parameters and synoptic (predictor variables) and climatology Parameter that it is necessary to have a vision of the future(Instant predictor variable) is a transfer function is provided. One of the applications that combines these two approaches based on statistical downscaling model (SDSM) is. The meteorological station data required as input and output in seven steps GCM model on the basis of daily data in the area are downscaled.
The third downscaling model is artificial neural network (ANN), developed by Coulibaly et al., 2005. This model is a non-linear regression type in which a relationship is developed between a few selected large-scale atmospheric predictors and basin scale meteorological predictands. In developing that relationship a time lagged recurrent network is used in which inputs are supplied through tap delay line and the network is trained using a variation of backpropagation algorithm (Principle et al., 2000). A slightly different approach is used in selecting predictors for the case of neural network downscaling.
To compare data generated models and observations can be compared to an average of two non-parametric test Mann-Whitney society that is using correlation analysis. For the observed data and the model can be generated from correlation Spearman used. The basic correlation analysis based on linear correlation coefficient of the two variables. One of the important indicators that can be used for performance evaluation model, index model mean square error (RMSE) is defined as follows:The area North West of Iran, which includes the provinces of East and West Azerbaijan, Ardebil, Zanjan and a part of Kurdistan is the geographical coordinates '30 ˚49 '07 ˚44 to the East and the '00 ˚36 to '50 ˚39 North, is located. To study the effects of climate change in the region, using statistical models mentioned the need for a minimum period of 1961-1990 is based. In addition to the complete statistical period synoptic meteorological stations of old climate data confirmed the country's Meteorological Agency has been helping though some regional stations are multi-year statistical vacuum.
The results show that according to the Mann-Whitney test the performance of three models for minimum temperature in the study area are close. Spearman correlation test results for minimum temperature show that the number of correlation, in all stations for LARS-WG model is less than the other two and demonstrate low performance LARS-WG model is in this respect. The average number of months with significant correlation for ANN model with seven months of the year, the best performance among the three models in this respect. SDSM model with a four-month correlation table in the middle. In terms of RMSE index for the minimum temperature, LARS-WG and ANN models have average values are close together and show the error of sum of squares closer together the two models. RMSE values are less than the SDSM model and this shows the SDSM model is less than the other two models.
According to our evaluation, according to Mann-Whitney test data generated in which the difference between the observed and tested model placed, Parameters for minimum and maximum temperatures, three models have not different performance. But the results were somewhat different in different stations. Correlation data for SDSM and ANN models for maximum high temperature and minimum temperature for solidarity in SDSM model is less than ANN model. However, because the same structure prediction methods and large-scale use of such an outcome was not unexpected.
Mann-Whitney test for precipitation results show that significant differences observed and modeled data for ANN model is much more than the other two, which reflects the low performance of this model. SDSM and LARS-WG model and have similar good performance in this regard. The Spearman correlation test, all three models have a low correlation was observed and the model and represents the three models in the study area in this respect is low. According to the RMSE, the SDSM model is better than the other two models have average performance.

Flood and drought have caused several damages in natural and unnatural ecosystems in recent decade. Rainfall prediction can be useful in water resource management. The goal of this study is modeling the monthly precipitation of south east of Iran in South-Baluchistan basin, using artificial neural network (ANN) and stochastic models. This area has an unpredictable and complicated monthly rainfall pattern due to impact of several different precipitation systems of other surrounding regions. SARIMA time series models and Time Delay Neural Network (TDNN) are used in monthly precipitation forecasting. Monthly time series of rainfall during 1351-52 to 1387-88 in selected station were used in this study. Stations selection was based on Geographical distribution and data quality. Comparing the results of models of forecasting showed that TDNN model is superior to SARIMA time series model due to different rainfall systems and very sporadic precipitation in this area.

This study performs a sensitivity analysis to evaluatethe meteorological parametersthat affect daily pan evaporation rate. To this end, five meteorological parameters namely, daily mean temperature, relative humidity, sunshine hours, solar radiation, wind speed and pressure for period of 1386 to 1390 were used at the Tabriz City, Iran. At first, the pan evaporation rate was estimated using Artificial Neural Network (ANN) and the best structure of the ANN was distinguished. Then, weight matrix of selected structure of the network along with the Garson algorithm were used for sensitivity analysis of the input parameters and determine relative importance of the input parameters. The results indicated that the daily mean temperature and relative humidityare the most effective variables. However, the sunshine hours, solar radiation, wind speed and pressure have less effect on the evaporation rate at the Tabriz station.

A deficit in precipitation (meteorological drought) can result in a recharge deficit، which in turn causes lowered surface flow and a deficit in groundwater discharge. Given the importance of water in human life، regulating the access to reliable and sustainable water resources and planning proper consumption are essential for every designated region. There are two type of limitations that results from natural phenomena or improper management by human. This phenomenon is evident when above mentioned two factors emerge together. The purpose of this study is to identifying the climatic conditions that affect the flow in Dehgolan basin. The applied dataset in this study is the Precipitation، temperature، evaporation and runoff recorded in stations located at the Dehgolan basin. First using the Double-Mass curve the accuracy and the exactness of the mentioned data checked. Having made sure of their accuracy، using the data of adjacent stations and through proportions and differentials، the lost data of each station rebuilt. Drought occurrence was calculated using SPI index and other climatic variables normalized too. Then operative climatic conditions on surface flow studied using the artificial neural network in MATLAB environment as the method of feed forward back propagation. The highest correlation coefficient and proper mean square error for the input parameters obtained in an input model include: SPI in half year time scale، flow in the last months، temperature and evaporation in the synchronic month. Compare the multiple regression method and artificial neural networks shows higher correlation coefficient in artificial neural network. According to the major changes in the values of correlation، Standard Precipitation Index (SPI) and the discharge of the previous month can mention that the variation of these parameters got a higher effect on decreasing or the increasing of the monthly discharge.

In this research, by review of previous works and field works, the Nine factors identifiedeffective factors in landslide hazard and used for analysis risk by GIS software. the occurredland slides in the study area were gathered and rectified by GPS. These Nine maps werecrossed with the occurred landslide map and Landslides amounts and their areas werecomputed in each class. After determining the rate of each factor (element), land slidezonation was performed in GIS by artificial neural network and AHP Models. The efficiencyof output results of models was assessed by DR and QS indices. The results of DR indexshowed The map was produced using a neural network than maps produced using the analytichierarchy higher accuracy for the study area.

Notice of the amount of the claim in the context of license of building in each course is one of the fundamental issues that municipals for answer to client needs. The lack of information in this area caused problems such as a waste of time and energy، reducing efficiency and dissatisfaction with clientele and ultimately causes of the lack of planning. With regard to process of non-liner and pendulous individuals motivated to building and more crafting of construction permits from municipalities and variable affecting it، non – liner models special Artificial Neural Network (ANN) and Adaptive Neural Fuzzy Inference System (ANFIS) would have more success. To this end، consider a combination of the most essential parameters as the external and internal part that are impact on the decision of people making to building، inclusive city''s population and its grows rate، average of the household costs and revenue، the impact of the season`s and temperature، the rate of gross domestic product (macro level)، inflation، and Exchange rate fluctuations (as the external parameters) and factors such as land and its price، building density، and the rate of imposition of construction (as the internal parameters). in continue to compare their ability to both، utilize from evaluation criteria to performance models such as the coefficient of determination (R2)، middle absolute deviations (MAD) and root middle square errors (RMSE). Finally ANFIS excel to ANN due to reliance to combination of ANN ventilations power and logical function of fuzzy inference system with amount of R=­­ (0. 9899، 0. 9656)، RMSE= (0. 0064، 0. 0026)، and AMD= (0. 0061، 0. 0018) for training and testing system.

Being available the accurate data on carried sediment has accounted as an important factor for making decision about constructing of river structures and determining of dam life. To accomplish this object، a number methods have been proposed so that sediment rate curving as an hydrological method has been developed for doing it. Ignoring differences among seasons causes to lower the precision of this method. So، present research has been programmed for evaluation of classified discharge to two categories including high water and low water on suspended sediment calculated by sediment rating curve in comparison with Artificial Neural Network (ANN). For acquiring this object، by means of flow duration curve and USBR method، daily suspended sediment and sediment rating curve were resulted. Finally، some statistical criteria including Relative Error (RE)، Model Efficiency (EF)، Root Mean Square Error (RMSE) and Descriptive Coefficient (R2) were applied for comparing the results outcome of sediment rating curve method and ANN method. Results showed that ANN method has as higher capability in comparison with sediment rating curve on basis of Descriptive Coefficient and Model Efficiency 0/903 and 0/89 respectively moreover Root Mean Square Error and Relative Error 0/322 and 6/22 respectively.