maryam riahinia

Today, due to population increase, industrial development, excessive exploitation, droughts, exploitation of underground water has multiplied. Therefore, identifying areas with underground water as one of the important sources for providing drinking water, agriculture, and various industries is considered to be one of the important and necessary issues in water resources management. The purpose of this research is to investigate and zonate the areas with underground water in Khorram Abad plain located in Lorestan province using convolutional neural network method. For this purpose, maps of nine factors affecting underground water were first prepared in the ArcGist environment. In the convolution method, the number of samples was determined as the ratio between the training set and the test set was 70:30, and the convolution neural network framework was used as 2 convolution layers and 2 integration layers, 2 complete connections. layers and finally the sigmoid layer was used for classification from the 3-3 convolution kernel, the Relu function as the activation function and the cross entropy function as the loss function. The obtained maps were classified into 5 classes: very good, good, average, low and very low. Confusion matrix was also used to validate the results of the model. 30% of the real data was used for evaluation, which resulted in an overall accuracy of 92%, that is, the model was able to correctly identify 92% of the data as underground water and 93% as the absence of underground water. The analysis of the groundwater potential map of the convolutional neural network model shows that about 57% of the area is in low groundwater conditions and 43% of the area is in good groundwater conditions.

In recent years, with the growing population, industrial development of groundwater resources has more than doubled, with groundwater levels continuing to fall and eventually reaching a point where there will be no more water to extract. Therefore identifying these resources, making optimum use of them means a permanent and permanent harvest of this natural gods wealth. 

In this first phase, field studies and data collection were done. Wells map was prepared first and then effective parameters were identified: elevation layer, slope, slope direction, surface curvature, topographic moisture index, land use, soil, geology, river distance, drainage density, fault distance, fault density specified. And their plan was made. Satellite images of ENVI and eCognation software were used for mapping the land and the images were classified with basic pixel and object crosshairs and fuzzy logic and artificial neural network methods were used. Based on fuzzy logic, baseline maps are first ranked based on their impact and importance on groundwater resource potential, and then determined using a fuzzy method for each specific class rating factor. In the neural network method, these agents, along with a number of wells, enter the network as the input layer. In this way, the pattern is trained by the network between the input parameters (network input) and the areas where potential water resources exist (network output), then for The input parameters of the catchment to the trained neural network are predicted corresponding outputs which are potential areas of groundwater potential. Then the results of both models are tested and finally by comparing the results of the neural network model with the fuzzy logic model an appropriate method for groundwater resource potential in the catchment is obtained and the most important factors in resource potential in the area are identified.

Evaluation of effective layers in the potential of groundwater resources: The results of the elevation factor showed that the highest percentage (33%) of the area with high potential is located at an altitude of 1700-2000 m, which is the average altitude of the region, and the results of the slope factor study show that the highest percentage of high and medium potential areas is on the slope 0-1 is located and the highest percentage of areas with low potential is on the slope of 60- 173 and the highest percentage of areas with high and medium potential to the northeast direction and the highest percentage of areas with low potential is in the southern direction. The results of the study of lithology class (formation) show that the highest percentage of areas with high and medium potential. The weak are in the class (reserves of old and new mountaineering terraces and conifers). And the highest percentage of soil in the region is in areas with high potential and weak in the group (rocky / intulse outlets) and the highest percentage of areas with medium potential in the group (rocky direction outlets / input solo).The results of the study of the distance from the river show that the highest percentage of areas with high potential is located in the nearest distance from the river 200 meters and the highest percentage of areas with medium and weak potential is in a range far from the river. The results of the fault gap survey show that the highest percentage of areas with high and medium potential is located in the closest distance from the fault (0-0.032) and the highest percentage of areas with low potential (0.115-0.170). Precipitation and temperature estimates show that the highest percentage of areas with high, medium and weak potential is in the range of 756-844 and the highest percentage of areas with high, medium and weak potential is in the range of 13-15. And the highest percentage of areas with high and low potential is in the use of medium rangeland and the highest percentage of areas with medium potential is in forest use. The results of the groundwater map show that the highest percentage (36.85) of areas with high potential is in the class of 10.12-6.74. Evaluation of the classification maps of the results shows that the accuracy of kappa in object-class classification is 96% and in base pixel 85%. Neural network results also show that about 39% of the area has high potential of groundwater, while fuzzy logic map shows about 87% of the area with low potential.

Using the descriptions needed for basic pixels and taps, we can provide you with faster access to various sites, loops, content and technical and engineering information in a variety of areas, in the entertainment and leisure markets. Read yourself below. From this research it can be concluded that your neural network can control the energy potential of water Underground is more practical than the appropriate fuzzy logic method because its results are closer to the ground. Factors such as slope factor can be considered as an important factor in the potential of groundwater resources because of the high percentage of potential areas in the range of 0-10%.

Much of Iran is dry and poor water, and the water required by villages, industries and cities is used by groundwater sources. Excessive groundwater harvesting in many parts of the world has caused a severe drop in groundwater levels. Therefore, this study was conducted to investigate the spatial and temporal variations of groundwater level in Khorramabad plain over a ten year statistical period (2004-2014) using the best method, geostatistical estimator. In this regard, firstly, the available groundwater statistical data were collected and various conventional and simple kriging and IDW (RBF) methods with Gaussian, exponential and circular models were used. The mediation method was performed using two criteria (ME, RMSE). Then their location maps were plotted in Arcmap 10.3. The results showed that the multi-quadratic photo-distance method (RBF) is the best groundwater interpolation method in 2004 and 2014.