Spatio-temporal monitoring of tea plantation phenology using RISAT-1A SAR and Sentinel-2 optical data

Tea is a vital perennial plantation crop in India, requiring continuous monitoring to support timely cultural operations and maintain yield quality. This study evaluates TIMESAT’s capability to detect key phenological and agronomic stages in tea plantations using Sentinel-2 optical time-series and assesses spatial variation in the onset and end of the picking season across the Palampur sub-district of Kangra, Himachal Pradesh, from 2020 to 2022. RISAT-1A SAR-derived surface scattering was correlated with field-measured Leaf Area Index (LAI), revealing a moderate relationship (R2 = 0.51), while NDVI showed a stronger correlation with LAI (r = 0.68) and was selected for TIMESAT-based processing. TIMESAT effectively captured picking and pruning cycles and revealed notable spatial variability in seasonal timings, enabling identification of gardens exhibiting delayed or irregular phenological behavior indicative of inadequate management or abandonment. Validation against manually interpreted stages showed a mean absolute error of 28 days and an RMSE of 34 days at the optimal threshold of 0.5, confirmed as most robust through sensitivity analysis. By shifting from manual interpretation to automated phenology extraction, this study provides a scalable framework for monitoring economically important tea growth stages, offering valuable insights for agronomic decision-making, labour planning, and supply chain coordination.