| Popis: |
Identification of crop area and its extent estimation is one of the important pre-requisite for timely decision making. Rice is a principal food crop and the extent of area sown depends on the availability of water resources. To study the response of direct sown and transplanted rice crop to dual polarized multi-temporal Sentinel-1A data and estimation of rice area by synergetic use of SAR and optical data was carried out in the Guntur district of Andhra Pradesh, India. The mean backscatter coefficient values range during the agronomic flooding in transplanted rice was recorded -23.1 dB for VH and -13.7 dB for VV bands. The increase in backscatter coefficients is more continuous at the VH polarization after the transplanting of transplanted rice and after flooding the fields in direct sown rice. A sharp decline in backscatter coefficient values was observed at the heading. Water availability is one of the major factor which deciding to choose the duration of the cultivar in the study region. Direct sown condition supports the long duration cultivars. The rice area was estimated using VV, VH and NDVI revealed that, an area of 2,25,239 ha of the rice crop was cultivated during the kharif 2020 crop season which was found to be overestimated by 5.2% as compared to the data of local administrative authorities. The overall accuracy of the estimated rice area map was 97.2 per cent with a kappa index of 0.95. KEYWORDS: Sentiel-1A, SAR, rice, multi-temporal and SVM algorithm |
