The researchers improved a hydrological model in the Upper Huai River Basin by using streamflow observations to update the model states. By assimilating the streamflow data into the model using the Ensemble Kalman Filter, they were able to significantly enhance the accuracy of streamflow predictions. Updating the model states with streamflow observations helped to better estimate the flow of water in the basin, especially at interior sites and near the catchment outlet. This approach showed promising results in both synthetic experiments and real data applications, demonstrating the potential of using observed streamflow data to improve hydrological modeling.