MAPPING SOFIA PLAIN ARABLE LAND DYNAMICS USING LANDSAT-8 OLI IMAGES AND GROUND DATA

The purpose of this article is mapping the arable land dynamics using satellite and ground data. The methodology of this article includes the following working stages: 1) Choosing appropriate satellite images; 2) applying arable land mask from CORINE 2006 land-cover database; 3) Deriving Normalized Difference Vegetation Index (NDVI) images and composing unsupervised classifications on the chosen satellite images; 4) Conducting per-pixel supervised classification using the Maximum Likelihood Classifier (MLC); 5) Applying accuracy assessment tool in ERDAS Imagine and deriving accuracy totals and Kappa statistics; 6) Mapping the arable land dynamics. The investigation shows good results for mapping arable land dynamics. The overall accuracy for the LANDSAT-8 Operational Land Imager (OLI) satellite image, acquired on 23.03.2014, is 77.66 % and Kappa statististics is 0.6906, while the image acquired on 14.08.2014 was estimated at 86.02 % overall accuracy and Kappa statistics of 0.7646. The analysis on the LANDSAT-8 image, acquired on 14.08.2014 shows that it can be used for controlling and monitoring the agricultural threathments accomplished by the farmers like harvest, for example. The presented research shows the big potencial using LANDSAT-8 OLI data for crop identification and mapping arable dynamics purposes at a relatively high accuracy.