High-resolution topographic datasets collected by light detection and ranging (LIDAR) sensors capture subtle elevation gradients influencing hydrologic processes across low-relief landscapes such as river floodplains.  By viewing a floodplain as a collection of interconnected and internally draining patches, we developed a novel GIS strategy to assess patterns of hydrologic impedance and map geomorphic features (topographic depressions and hydrologic divides) that determine surface water connectivity across river floodplains at any desired scale of resolution. The algorithm identifies fine-scale patches associated with depressions and calculates hydrologic impedance of divides among adjacent patches.  Coarser-scale patches can be derived by preferentially merging the fine-scale patches across the least important divides (i.e, those with the lowest hydrologic impedance).  For a section of the Umatilla River Floodplain (Oregon, USA), we demonstrate how this technique can generate representations of surface-water flow networks at multiple scales of resolution by identifying and representing the most influential geomorphic features controlling hydrologic connectivity at any desired spatial scale.  This approach provides an useful strategy for objectively quantifying patterns of hydrologic connectivity across low-relief landscapes (e.g., identifying hydrologic “nexuses” between rivers and wetlands as required under the U.S. Clean Water Act) or for generating optimized link-and-node flow networks to support hydrologic modeling.