جستجوی مقالات مرتبط با کلیدواژه "cloud computing" در نشریات گروه "آب و خاک"

Lack of water, increase in production costs, change in land use, and increase in demand for food have caused the accurate determination of spatial and temporal monitoring of the paddy fields area to be extremely important for planners and decision makers. Actually, the use of field methods to estimate the cultivated area of crops in large areas is costly and time-consuming, and is associated with errors. Hence, the purpose of this study was to use remote sensing images to estimate the paddy fields area in the Guilan Province by the best classification method. Therefore, Sentinel 2 satellite images were analyzed using 6 supervised classification methods including ML, CART, RF, SVM, GME and RF-NDVI methods in GEE environment. The ML method was performed in the ENVI environment and the rest of the methods were performed in the cloud space of the GEE environment. The results of using the classification methods showed that the Random Forest method along with the NDVI (RF-NDVI) with a kappa coefficient of 0.94 and total accuracy of 0.90 had the highest accuracy compared to the other methods, which shows the effect of the vegetation index in distinguishing between paddy fields and other land uses. This estimation of the area under rice cultivation in the province showed that the net area of the total paddy fields in the province was 218,135 ha. This estimate, compared to the available statistics of the Agricultural Jihad Organization (238,012 ha) and the Regional Water Company of Gilan Province (245,000 ha) was, respectively, 8.35%. and 10.96% less