The use of fine-resolution LiDAR DEMs is becoming increasingly commonplace in environmental modelling, and particularly in hydrological modelling. However, the incomplete representation of transportation embankments and their associated drainage infrastructure in fine-resolution LiDAR DEMs of heavily altered landscapes negatively impacts the accuracy of derived terrain and stream network attributes. This research investigates and develops a DEM preprocessing solution to remove the artefact dams caused by the presence of embankments at sites of underpass locations (e.g., culverts, bridges) and enforce flow through subsurface drainage infrastructure and roadside drainage ditches in fine-resolution LiDAR DEMs. Specifically, a novel modified least-cost depression breaching technique was developed to enforce flowpaths through transportation embankments at locations of automatically identified and/or mapped underpasses. The novel breaching technique reduced the number of erroneous breaches across transportation embankments by an average of 92.5% compared to its predecessor algorithm, using commonly available ancillary data.