USING REMOTE SENSING IN DETECTING SUGAR BEET FIELDS TREATED WITH DIFFERENT DOSES OF PHOSPHORUS

Abstract

Nowadays, the use of remote sensing technology in agricultural applications is increasing rapidly. It allows determining the amount of the fertilizer required in an agricultural field and detecting fertilizer requirements, which are among the most significant factors affecting plant development, and its terrestrial components. Depending on the amount and type of the fertilizer used, the spectral signatures of plant leaves can be evaluated by remote sensing data. In the present study, it was aimed to determine the impact of the difference in the phosphorus amount used in the sugar beet plant on the reflectance of the plant. In this context, 3 different areas were identified in the study field. Sugar beet was planted in these areas by applying different doses ranging from 0 kg P ha(-1), 150 kg P ha(-1) to 300 kg P ha(-1). Spectroradiometer measurements were carried out simultaneously with Landsat 8 satellite transition (satellite image dated 17.09.2017) for obtaining the spectral signatures of the sugar beet treated with different doses of phosphorus when the plants were grown to 5 months old. The Matched Filtering (MF) algorithm was used as a spectral classification in which spectroradiometer measurements of the sugar beet were used as a reference endmember in order to detect different phosphorus groups. Furthermore, the Normalized Difference Vegetation Index (NDVI) was applied to the satellite image to compare the MF result. According to the results of MF and the NDVI which are coherent with each other, it was concluded that sugar beets in three different dosing areas were correctly identified in different classes.