Nested multi-source data assimilation coupled with a hydrodynamic model for enhancing water-quantity monitoring accuracy

Introduction Multi-source data assimilation combines observations and model simulations to improve water-quantity monitoring. Monitoring in the lower Lancang River is challenging because of complex basin conditions and hydropower regulation. This study aims to improve monitoring accuracy and efficiency. Methods We developed a nested multi-source data assimilation framework coupled with a hydrodynamic model. The framework integrates manual measurements, side-scan radar observations, and hydrodynamic simulations. A LASSO-based enhancement scheme was used to improve side-scan radar estimates, while the nested assimilation system extended point-based observations to the river reach. Results At Yunjinghong Hydrological Station, the LASSO-based scheme improved the accuracy of the side-scan radar monitoring system by 22.93% compared with conventional methods. The nested assimilation framework achieved a correlation coefficient of 0.935 during the validation period. It also provided water level, discharge, and flow velocity at locations throughout the modeled channel. Discussion The framework offers a technical approach for improving water-quantity monitoring in complex, regulated rivers. By extending station observations across a river reach, it may support water-resources management and cooperation in the Lancang–Mekong Basin.