Quantifying hydrologic exchange fluxes (HEFs) at the stream-groundwater interface and their residence time distributions (RTDs) in the subsurface are important for managing the water quality and ecosystem health in dynamic river corridors. However, field measurements and directly simulating high-spatial resolution HEFs and RTDs can be time-consuming, particularly at watershed-scale. Recent research has proposed the potential application of using hydromorphic units (HUs), which cluster the bathymetry and surface water hydrodynamics attributes, to aid in RTD estimation. However, more pioneering studies have demonstrated that underground structure and large-scale river channel geomorphology are the primary factors controlling the HEFs and RTDs. To address this contradiction, this work evaluates the HEFs and resulting HU-RTD relationships for two 10-km long river sections along the Columbia River, leveraging a one-way coupled three-dimensional transient surface–subsurface water transport modeling framework. Applying such a framework at the two river sections with similar HUs allows for quantitative comparisons of HEFs and RTDs using both statistical tests and machine learning classification models. The comparison reveals that the similarity and transferability of the HU-RTD relationship is very low for the two investigated river sections. This suggests that, creating a general algorithm to estimate RTDs in large-scale river reaches based solely on surface water hydrodynamics and short-distance bathymetry topography data may be nearly impossible.