<p>Floods are known to cause extensive damage to property and life, which makes it necessary to determine the plausible magnitude and frequency of these hydrologic extremes. In this study, we use a combined hydrologic and hydraulic modelling approach to study the flood characteristics of Brahmaputra - a large transboundary river (580,000 km<sup>2</sup>) associated with complex topography, geomorphology, and a dense network of tributaries. A semi-distributed process-based model HEC-HMS, (forced with different precipitation datasets, -(APHRODITE, GLDAS, IMD and TRMM) is used to simulate an ensemble of its historical streamflow at a daily timescale. These calibrated and validated flows are used in conjunction with a network of level gauge stations (within Arunachal Pradesh & Assam in India) to quantify and improve flood mapping with the help of a two-dimensional hydraulic model HEC-RAS. The historical flood extents (2015 to 2022) obtained using the hydrologic & hydraulic modelling approach is further validated with the help of satellite earth observation data products. Such a multi-pronged, ensemble-based modelling strategy has the potential to create a more informed flood risk management system, in terms of providing better likelihood and uncertainty estimates, in the lower reaches of Brahmaputra often prone to prolonged inundation.</p>