Identification of Rice Cropping Season and High-Resolution Area Estimation in Major Rice Tracts of Kerala Using Sentinel-1 SAR and Sentinel-2 Optical Imageries

Abstract

Accurate identification of rice cropping seasons and precise estimation of cultivated area are critical for crop monitoring, water allocation, and greenhouse gas modeling. In this study information on  seasonal vegetation dynamics derived from monthly VH backscatter (σ0) from Sentinel-1 C-band Synthetic Aperture Radar (SAR) time-series (VH polarization)  were cross-validated with Sentinel-2 NDVI time series and field surveys during 2023-24 and 2024-25, enabling clear detection of Puncha season (November–March) and Viruppu (May–September) in Kuttanad and Mundakkan (November–March) in Palakkad which are the major paddy cultivation tracts of Kerala. Supervised Maximum Likelihood Classification (MLC) of Sentinel-2 imagery, supported by VH and NDVI time-series, estimated 46,960.65 ha of rice in Kuttanad and 20,519.50 ha in Palakkad, with Kappa coefficients of 0.87 and 0.83, respectively. This research focused on block-level mapping and generation of rice concentration information, which revealed intra-district variability and hotspots of rice cultivation. This confirmed the robustness of SAR-based monitoring for crop season detection and use of Sentinel-2 for high resolution rice area estimation in fragmented, cloud-prone tropical landscapes.