APPLICATION OF SENTINEL-1 MULTI-TEMPORAL DATA FOR CROP MONITORING AND MAPPING
- 1Dept. of Geomatic Engineering, Faculty of Civil Engineering, Yildiz Technical University, 34220 Esenler, Istanbul, Turkey
- 2Dept. of Geology, Faculty of Mines, Istanbul Technical University, 34469 Maslak, Istanbul, Turkey
Keywords: Sentinel-1, SAR, Backscatter, Time-series Analysis, Agricultural Monitoring, Crop Mapping
Abstract. The Synthetic Aperture Radar (SAR) technique superiority has provided various opportunities in agricultural studies mainly on crop monitoring and management. This has resulted in many different investigations and improvements in crop growth monitoring, crop disaster prediction and providing accurate information to precise farming. In this study, a time-series of Sentinel-1 SAR images were acquired throughout the agricultural season synchronously with the field measurements to investigate the temporal backscatter changes for different crop types. From the backscattering analysis, it was observed that each similar crop type in different test fields due to the distinct methods of irrigation and fertilization has shown different intensity values. However, they all follow a roughly identical backscattering trend during growing stages of the crop and useful information can be extracted, such as estimating irrigation and harvesting time according to the changes made in backscatters. In comparing with ascending orbit in VH and VV polarizations it has indicated that the homogeneity between SAR backscatters is high for each field with the same crop type in descending pass direction with VH polarization. In contrary, high-intensity values are recorded in VV polarization for entire crop types. It is also observed that polarimetric composite images for a different date are useful to roughly identify crop types, and validated with the application of classification methods in the study area. As preliminary results, it is demonstrated that SAR data provide useful information about crops status. Hence, Sentinel-1 SAR data are ideal preference due to its free availability and constant long-term data archive.