Rating curve estimation in ungauged basins using coupled hydrological–hydraulic modelling and multi-source remote sensing and UAS data

Discharge observations mainly rely on gauged water levels through rating curves (RC), whose reliable establishment requires long term measurements that are often unavailable due to high maintenance costs, complex terrain, and political reasons. As a result, many basins worldwide remain ungauged, making RC estimation particularly challenging. Recent advances in remote sensing, including satellite altimetry, provide new opportunities for discharge and RC estimation in ungauged basins. However, several challenges remain, including parameter equifinality in discharge inversion from water level, oversimplified assumptions of channel resistance and cross-sectional instability in morphologically active rivers. While Unmanned Aerial Systems (UAS) enable retrieval of channel geometry in complex and hard-to-reach river reaches which imposes an important constraint to mitigate parameter equifinality in hydrodynamic modeling, a systematic assessment of how UAS and remote sensing observations can be combined to reliably estimate rating curves in fully ungauged basins remains lacking.

Funded by European Union's Horizon Europe project UAWOS (Unoccupied Airborne Water Observing System), this study presents a RC estimation framework specifically for ungauged basins using multisource remote sensing, UAS data, and a coupled lumped rainfall–runoff and one-dimensional hydrodynamic model. The model is fully forced and calibrated using remote sensing and UAS observations only. To address parameter equifinality, we first perform a temporal-scale dependent parameter sensitivity analysis to assess the identifiability of model parameters given availability of different remote sensing observations. Based on the sensitivity results, a multi-staged Bayesian calibration strategy is introduced, in which each observation type constrains only the parameter subspace supported by its information content. Isar River, Germany was chosen to test and evaluate the feasibility of the proposed methodology.

Overall, the proposed framework provides a transferable theoretical and technical pathway for estimating RC in ungauged river basins, demonstrating the potential of combining UAS and remote sensing data to derive RC, without relying on prior discharge measurements and offering implications for estimation of ungauged catchments.