جستجوی مقالات مرتبط با کلیدواژه "data mining" در نشریات گروه "آبخیزداری، بیابان، محیط زیست، مرتع"

Although land degradation is a worldwide challenge and a destructive phenomenon, little studies have been done on the application of new numerical methods (data mining and statistically), for spatial simulation of this phenomenon and identification of areas sensitive to land degradation. The aim of this study is to spatially simulate land degradation in the Qazvin plain using the frequency ratio model to identify areas prone to land degradation. For this purpose, using the trend of changes in net primary production during the years 2001 to 2020, the points of occurrence of land degradation in the Qazvin plain were determined. Approximately 70% and 30% of the points were used to prepare the land degradation vulnerability map and validate the model's efficiency, respectively. For this research, 15 parameters affecting land degradation (directly and indirectly) including temperature, rainfall, slope, aspect, elevation, EC and SAR of ground water, ground water level, annual ground water decline, land use, normalized difference vegetation index, normalize difference salinity index, vegetation soil salinity index, normalized difference moisture index, and visible and shortwave infrared drought index, were introduced into the model as predictors factors or independent parameters. Finally, using the area under the ROC curve, the effectiveness of the frequency ratio model for spatial simulation of land degradation was assessed. The map of land degradation susceptibility shows that the areas prone to degradation are located in the northeast, north, northwest, west, southwest, and south of the Qazvin plain, which mainly includes good, moderate and poor rangelands. For the land use parameter, the highest frequency ratio was associated with the sum of good, moderate, and poor rangeland (5.66). The value of AUC = 0.7 indicates the good performance of the frequency ratio model in spatial simulation of land degradation.

Today, the use of data recorded in synoptic stations of the country is one of the most significant sources of applied research for researchers. Data recorded automatically or manually at synoptic, climatological, and other stations are analyzed for statistical analysis. In this research, the data recorded in the synoptic stations of Iran, which are used to determine the days of dust, were analyzed using the science of monitoring and data analysis using ensemble classification. In this study, data from 36 synoptic stations, were used. These stations are in Isfahan, Kerman, Yazd, Sistan and Baluchestan, Semnan, Markazi, Khorasan Razavi, Hamedan, Qom, and South Khorasan. The parameters of daily average temperature, daily rainfall, station height, geographical location of the station, maximum wind speed, maximum wind speed, and dew point were used for the classification. The results showed that the most important factor among these parameters for dust is the maximum wind speed, which was identified as the most significant factor in all classification methods. Also, three classifiers, KNN, SVM with RBF kernel, and MLP neural network, were selected as members of the ensemble, which accurately detects 90.7 percent of the source of dust production (from the inside and outside the dust).

Identification of climatic characteristics affecting the annual yield of Rheum Ribes can be useful in management and development of this species in the rangelands. In this research, the annual yield of this species in Khorasan-Razavi province based on 74 climatic parameters during a ten-year period evaluated and affecting climatic parameters extracted using data mining methods. First, the role of climatic parameters associated with temperature, humidity, rainfall and sunny hours analyzed using correlation and regression methods. Then, 11 classification algorithms in MATLAB software programmed and compared. The results showed that the Rheum Ribes yield has a positive relationship with the average of maximum temperatures in the summer, the range of high temperature in May to September, the maximum of summer temperatures and the relative humidity and rainfall of the spring. Evaluation of the algorithms using the indices of coefficient of determination and mean square error showed that in estimation of the annual yield based on climatic factors, the pattern recognition method at the testing stage with a coefficient of determination equal to 0.46 and regression methods, classification discrimination and K nearest neighbor (KNN) at the training stage (coefficient of determination equal to 1) had the best performance. With regard to the effective factors in stepwise method, the linear regression method at the testing stage (coefficient of determination equal to 0.74) and K nearest neighbor method at the training stage with coefficient of determination equal to 1, estimate the Rheum Ribes yield more accurately. Also, the proposed K nearest to mean (KNM) method for k values equal to 6 and 7 with regard to all factors and the effective factors resulted from stepwise method, respectively, had higher accuracy in yield estimation. So, application of data mining methods and the proposed model, in recognition of climate parameters affecting different rangeland species could be a practical approach.

Conventional soil mapping is related to High density sampling, affected by scale and expert knowledge So using of new data mining methods in digital soil properties mapping was the main aim of this study for resolving conventional soil survey problems. In this research, 62 surface soil samples based on regular grid and expert knowledge opinion were selected after that soil organic carbon(SOC), clay content and CaCO3 were determined in some part of Dryland Kuhin region with area of 372 ha. Data sets were divided to two 80%(calibration) and 20%(validation), respectively. From digital elevation model with 10-meter spatial resolution were derived 19 geomorphometric attribute in SAGA GIS software. Three geomorphometric covariate included TPI, TRI, DEM and landform map unit were chosen PCA and expert knowledge. RStudio and SoLIM Solution software were used for random forest (RF) and fuzzy logic modelling, respectively. The RF modelling results show that for SOC, clay and CaCO3 based on determination coefficient (R2) had 0.63,0.75,0.63 and RMSE 0.17,7.5,5.77 percentage and for SoLIM method revealed that R2 0.47,0.42,0.42 and RMSE 0.2,8.08,4.68 percentage, respectively. Generally, the RF model with creating nonlinear relationship among soil properties and environmental covariate can predicted digital map with appropriate precision for management and sustainable land utilization

Due to the lack of information in most of the watersheds, many researchers attempt to use spatial analysis within Geographic Information System (GIS) in hydrological and Flood Prone (FP) area studies. The present study was designed to compare the efficiency of three models i.e. Support Vector Machine (SVM), Generalized Linear Model (GLM) and Generalized Additive Model (GAM) for preparing the flood susceptibility mapping in Guilan province, Iran. For this purpose, slope, aspect, plan curvature, elevation, distance from the river, drainage density, geology, land use, Topographic Wetness Index (TWI) and Stream Power Index (SPI) layers were derived in GIS (ArcGIS and SAGA-GIS). Using 220 flood locations, 70% and 30% out of total flood locations were then used to calibrate and to validate the performance of the models, respectively. The evaluation results of the models accuracy using the area under the curve (AUC) and Kappa indices showed that in terms of AUC, the SVM with 0.835 and the GAM with 0.827, and the GLM with of 0.79 performed very good and good classes, respectively. In terms of Kappa index, the SVM with 0.58, GAM with 0.53 and GLM with 0.48 are performed good and acceptable classes, respectively. Therefore, based on the mentioned indices, the SVM superior to other two models for identifying the flood susceptibility areas.

Increasing population and agricultural development need dramatically water resources groundwater resources, therefore, are increasingly being considered, especially in arid and semi-arid regions. Aim of this research is mapping potential of groundwater resources on Yasouj-Sisakht region using data mining method Random Forest (RF) and Generalized Linear Statistical Model (GLM). For this purpose. For this purpose, information layers including slope, slope direction, slope length, aspect, topographic wetness index (TWI), distance from fault, distance from the stream, rainfall, land use, lithology, topographic position index (TPI) and stream power index (SPI) as the main factors influencing groundwater potential were identified and developed in ArcGIS and SAGAGIS software. From the distribution of 263 springs in the area, 70% (253 springs) were used as educational springs and 30% (109 springs) were used as experimental springs. The results showed that the level of underground water with low, medium, high and very high potential in the map of the random forest was 37.78, 22.22, 18.89 and 21.11%, respectively, and in the generalization linear model were 14.49, 32.04, 31.11 and 22.36%, respectively. Moreover, Sensitivity Analysis show that the factors affecting both methods are rainfall, altitude and distance from the fault factors. The accuracy of the data mining models used in this research was also evaluated using a relative performance curve (ROC). The area under curve (AUC) for both RF and GLM models is 92% and 65%, respectively. The accuracy of RF model, therefore, mapping groundwater potential in the study area is more than GLM model.

One of the important relations in the erosion of gullies is to study the threshold of erosion creation and expansion. In recent decade, creation of new knowledge in determination of relation between variables was led to develop prediction methods in different science and therefore, investigating the ability to use these methods in erosion and soil conservation is essential. Also, in order to control the erosion of the gully, the mechanism of gullies growth and its dimension expansion, especially increasing in gullies length, has to be carefully determine; for this purpose, the present study aimed to determine the threshold of the most effective factors on increasing the length of the gully, using the K-Means data mining algorithms and the CART decision tree in the Ghazian watershed in the north of Fars province. The results of this study, which include measuring various variables of gullies under field condition and in laboratory, and using data mining techniques, showed that increasing the length of gully in this area depended on the factors of the area above headcut, saturated extract electrical conductivity, forehead slope, canopy cover percentage, and sodium adsorption ratio. It is recommended control of erosion in the foreheads is highly important in reducing the increase in gullies length and sediment production. Also, improving the soils of this area with soil amendments and the restoration of compatible vegetation and the increase in soil organic matter should be considered as the priority of effective actions to control the increasing length of gullies.

Nowadays, water quality is one of the key issues in most regions all over the world. Because of its high relation with human health and also have important role in management and exploitation of resources. This research has done first time in Pasargad city located in 105 km of eastern north of Shiraz. In order to determine drinking water quality factors using data mining technique, results of this research have obtained using Quest decision tree in Clementine software(12 version) and showed most effective factors of drinking water in this region is function of total harness and electrical conductivity. In this way, if total hardness in this city is less than 282.232 parts in million and electrical conduction is less than 822.787 micro mohs in centimeter, this water will be suitable for drinking. Therefore recommended more noting to refine and decrease the hardness of water for human drinking water, moreover continuous refining in periodically regular sampling of water resources in this city should be noted.

Understanding of suspended sediment rate is one of the fundamental problems in water projects which water engineers consistently have involved with it. Wrong estimations in sediment transport cause incorrect design and destruction of hydraulic systems. Due to the difficulty of suspended sediment measurements, sediment rating curves is considered as the most common method for estimating the suspended sediment load. The main purpose of this research is the capability challenge of this method in comparison to some state of the art models. In this study, we selected some computational intelligence models (i.e. K-nearest neighbor (KNN), artificial neural networks (ANN), Gaussian processes (GP), decision trees of M5, support vector machine (SVM) and evolutionary support vector machine (ESVM)) and compared them with their sediment rating model in 8 basins located in Gilan province. Daily sediment and discharge data considered as the input data for 30-years. Evaluation of the results indicated that the Gaussian process model has the lowest residual sum of squares (RMSE) and the highest correlation coefficient (r) than the other models.

Estimation of suspended sediment load is one of the most important and fundamental challenges in the studies of sediment transport and river engineering, due to the damage caused by the lack of attention and considering it. Given the importance and role of sediment in the design and maintenance of hydraulic structures such as dams and As well as planning for efficient use of downstream of river and also conservation of nutrients at the upstream of river always lots of efforts have been done in the field of suspended sediment load estimation and numerical methods have been developed in this case. But due to the cost of most procedures or lack of adequate accuracy in most of common experimental methods, need to a new method that can estimate suspended sediment load with the greatest possible precision, seems to be very necessary. In this study the amounts of suspended sediment loads have been estimated with support vector regression and k-Nearest neighbor methods. Results indicated the acceptable ability of both data mining techniques that explored in this study in estimation of suspended sediment load. Among the methods examined in this study, the support vector regression method estimated the amounts of suspended sediment load in Lighvan Chay River with representing evaluation indexes such as (CC=0.959, RMSE=43.547(ton/day)) is more accurate rather than K-nearest neighbor method.

Data mining enables generalization of data of soil to remote areas and which is able to up/down scale of data in wide ranges of level that facilitate the decision-making process of executives. Cation Exchange Capacity (CEC) is one of the most important parameters in soil database and shows the ability of a soil to retention of minerals and pollutants. Due to low organic matter and specific mineralogy of soils in arid and semi-arid regions, measurement of CEC is time consuming and expensive. The objective of this study was to evaluate Coactive Neuro-Fuzzy Inference System (CANFIS) in prediction of CEC in soils of arid and semi-arid regions. A total of 85 soil samples from target area were selected among 440 soil sample database (available reference database) with a ratio of 1:5. Correlation test was conducted to assess the co-linearity of independent variables. Forward regression model was used to determine the most important and influential input parameters on the output results. The results indicated the reliability and high performance of the CANFIS approach in estimation of CEC using easily measurable characteristics, organic material, and satellite images.

The correct estimation of river discharge is an important issue in forecasting of drought and floods, designing of water structures, dam reservoir operation and sediment control. For this reason, water resources managers used intelligent techniques such as Artificial Neural Networks and data mining methods such as Decision Tree to reliably estimate the discharge in a river. In this study, the Elman Neural Networks (ENN) and M5 model trees were used to forecast daily discharge of Aharchay River. The daily discharge data of Aharchay River measured at the Orange hydrometric station was used for modeling. The results showed that for the forecasting discharge of one day ahead, the ENN method presents more accurate results in compression with M5 model. For forecasting discharge of one day ahead, the best scenario of ENN model with a relatively complicated structure of 9-3-1 that indicating 9 nodes in input layer, 3 nodes in hidden layer and 1 node in output layer, the calculated error measures were R2=0.90, RMSE=0.028 (m3/s) and MAE=0.001 (m3/s). The corresponding values for M5 model with only two input parameters including the discharge of current and last day, were R2=0.83, RMSE=0.734 (m3/s) and MAE=0.317 (m3/s). Comparing the performance of ENN and M5 models indicated that, however the ENN approach may present more accurate results than the M5 model tree, but the M5 model provides more understandable, applicable and simple linear relation in forecasting daily discharge. In addition, the number of required input parameter for M5 model is less than ENN model. Thus, the M5 model tree can be used as an alternative method in forecasting daily discharge.

Floods are the main natural disasters that produce serious agricultural, environmental, and socioeconomical damages in many parts of the world. Accurate estimation of river flow in streams can have a significant role in water resources management and in protection from possible damages. This study aims to compare the abilities of Support Vector Machine (SVM), M5 model trees and Linear Regression (LR) methods in forecasting hourly discharge flow of Aharchay River. The hourly water level-discharge and 14 flood events data of Aharchay River measured at the Tazekand hydrometric station was used for modeling. The results showed that the SVM method gives more accurate results than the M5 model and LR method in forecasting river flow for next one and two hours with the R2=0.96 and RMSE=0.0472 (m3s-1) and the R2=0.90 and RMSE=0.1596 (m3s-1), respectively. Comparing the performance of SVR and M5 models indicated that, however the SVR approach may present more accurate results than the M5 model tree, but the M5 model provides more understandable, applicable and simple linear relation in forecasting hourly discharge. Thus, the M5 model tree can be used as an alternative method in forecasting hourly discharge.

One of the most important types of water erosion is gully erosion. This type of erosion causes remarkable changes in lands، production of abundant sediment and deleterious social and economic consequences; especially in arid and semi arid regions of the world. Threshold is the point after which the behavior of the system changes and it can be related to the internal or the external factors of the system. One of the important questions about gullies is that in what condition the threshold of gully is created and expanded. Identifying the conditions seems important considering the fact that it is possible to compare conditions of the threshold with the regional conditions and determine how much the conditions are close to thresholds of gully creation and or Expansion. For this purpose، the current research aims to determine threshold of effective factors on length expansion of gullies in Mahourmilati region in south-west of Fars province by using the data mining techniques. The findings of the current research which were achieved by modeling based on K-Means Clustering Algorithms and CART Decision Tree and in Clementine software (version: 12. 0) indicated that the length expansion of gullies in the studied region is the function of SAR، soil acidity (pH)، and output area. To improve the soil of the studied region، it is recommended to consider improvements، revegetation and enhancement of soil organic matters prior to effective actions in controlling the length expansion of the gullies.

One of the effects of climate system change is the occurrence and intensification of drought phenomenon. Prediction of drought condition can play an important role in mitigation of its effects as well as effective management of the available water during the drought periods. Different approaches have been presented for evaluation of drought. Analysis of precipitation data is the general method for drought evaluation, as acceptable prediction of precipitation before its occurrence, would be necessary and effective for analysis of drought. The purpose of this research is the evaluation of the effect of data processing on applicability of two data mining models on drought prediction in Yazd station. In addition, during the recent decades some new computer based models have been developed for drought prediction and in most of the cases they have presented quite satisfactory results. In this research, prediction of precipitation, which is the main component on drought occurrence, has been carried out in Yazd synoptic meteorological station. Therefore, two data mining methods including Regression Decision Tree (RDT) and Artificial Neural Networks (ANN) have been used, and simulations were carried out in two different conditions. In the first condition, the measured values of some meteorological variables were used as inputs and the amount of precipitation was predicted 12 months in advance. However, In the second condition, 3-year moving average of data were the inputs of the models for prediction of precipitation amount 12 months before its occurrence. Finally for evaluation of the model performance in different conditions, statistical criterion including R and RMSE were employed. Results indicated that using moving average of data as inputs of the models has considerably improved the performance of the models. Both RDT and ANN methods are able to predict the amount of precipitation in Yazd station 12 months before its occurrence.