Reliable information of the amount of carbon sequestration of Indian forests at regional as well as national levels is necessary for providing information to support intergovernmental policy initiatives such as Intergovernmental Panel on Climate Change (IPCC). In our research, we focus on estimation and mapping of carbon stock and corresponding uncertainty over three heterogeneous forest ecosystems in India: Sholayar, Mudumalai and Araku – using a combination of sampled Global Ecosystem Dynamics Investigation (GEDI) data and wall-to-wall Sentinel-2 data within the Hierarchical Model Based (HMB) inference. HMB is a novel inferential framework for environmental surveys utilizing combinations of various sources of remote sensing and field data. In our research, we apply two modelling steps: the GEDI L4A model linking aboveground biomass (AGB) and GEDI data, and the Sentinel-based model linking the GEDI L4A footprint-level AGB density product and Sentinel-2 data. The proposed methodology using a combination of GEDI and Sentinel-2 data within HMB inference for environmental surveys complemented with statistically rigorous assessment of uncertainty will be a benchmark for carbon stock estimation across various forest ecosystems in India, where availability of field data is sparse. Our results will make a significant contribution to IPCC reporting, providing reliable information for intergovernmental policy initiatives.