Analyzing rice farming between sowing and harvest time with Sentinel-1 SAR data

Abstract

Agriculture has always been in an important position throughout human history. Today, the development of technology has accelerated studies to increase productivity in agriculture. With the use of remote sensing in agriculture, different crop types in large regions could be observed and their differences from each other could be examined with a spectral sight. With the observations obtained, instant surface monitoring in the agricultural sector makes it possible to perform analyzes. In the study, the paddy fields, where the rice product was named at the time of first planting, were examined by remote sensing method. Differences in Synthetic Aperture Radar (SAR) observations were analyzed between the first crop sowing and harvest time. In addition, in order to check the consistency of the results, the differences in the values obtained according to the representation of the samples distributed in the field were determined. Considering the results, it was seen that the lowest backscatter values were obtained for the paddy fields in the 35-day period after the first planting time and these values increased as the harvest time approached. There is an approximately 69% change in the lowest and highest mean backscatter values. Finally, when the time series analysis is performed according to the control samples in the field, it has been determined that the points represented by a single pixel have a more irregular distribution comparing the samples obtained in the form of polygons. This shows that pixels cannot be evaluated independently due to noise in SAR data.