RT - Journal Article T1 - Evaluation of OLI Sensor Data, The Capabilities of ALTA ReflectanceSpectrometerand Using The Concept of Virtual StationsMapping The Distribution of Heavy Metals in Soil JF - kntu-jgit YR - 2018 JO - kntu-jgit VO - 5 IS - 4 UR - http://jgit.kntu.ac.ir/article-1-547-en.html SP - 113 EP - 145 K1 - Remote sensing of soil K1 - Heavy metals of soil K1 - Virtual Station K1 - ALTA Reflectance Spectrometer AB - The use of remote sensing and GIS technology, is to have a quick and inexpensive method of obtaining information, especially in the field of environmental studies and earth sciences for optimal management. This study was carried out to evaluate OLI sensor data, the capability of Alta reflectance spectrometer in the selected spectral band satellite images for predictive models and the use of concept of station virtual changing the accuracy of the models and mapping the distribution of the concentration of heavy metals nickel, lead and zinc the surface soil around the Rajaee Power Plant. After determining the location of sampling stations on satellite images, according to the dominant land use around power plants (agriculture) and with respect to the different soil depth in this kind of usage,from the average depth of about 0-15 cm topsoil , 34 composite samples were collected and spectral readings were recorded using Alta reflectance spectrometer. The heavy metals concentrattion was determined in the laboratory using atomic spectrometer. According to the equivalent range of spectral bands of satellite imagery and spectral range light colored Alta reflectance spectrometer, was used for spectral data recorded by reflectance spectrometer to select the optimal band satellite imagery to participate the relationship between spectral data (independent variable) and the concentration of heavy metals in the soil (the dependent variable).The concept of virtual stations which are characteristics like spectral reflectance sampling stations are used to enhance the accuracy of the models. For this purpose, the OLI images, Landsat 8 satellite images, in the range of values of the pixels in all the bands (6 bands in the visible and infrared range), amounting 4450 virtual station was found in the area around the power plant. At the end using satellite image bands and finding the relationships between larger coefficient of determination (R^2), the distribution map of soil concentrations of heavy metals has been found with moderate accuracy. Validation of the models with the use of Root Mean Square Error (RMSE) showed the efficiency of use of the concept of virtual stations to increase the accuracy of the models; So that was observed in the linear regression, nickel, 17.21%, 20.19% lead and 4.05% zinc and nonlinear regression, nickel, 13.42%, 20.19% lead and 3.10% zinc increased accuracy in prediction models. It is suggested that in future studies data collected from other sensors should be evaluated as well. Other variables also affect the soil environment (humidity, temperature, vegetation, etc.) which should be considered in models. LA eng UL http://jgit.kntu.ac.ir/article-1-547-en.html M3 10.29252/jgit.5.4.113 ER -