HOW DO FLUVIAL NETWORKS BECOME RE-ESTABLISHED FOLLOWING GLACIATION?

In the formerly glaciated Central Lowlands, fluvial drainage networks were profoundly altered by ice lobes at the margin of the Laurentide Ice Sheet. During multiple glaciations, rivers experienced disturbances including diversion, reversal, incision, and complete filling of valleys with sediment. During interglacials, drainage networks evolved to integrate drainage of uplands. Drainage density increases systematically with time since most recent glaciation. Growth of fluvial networks apparently occurred across even very low relief areas, suggesting that river discharge was a major driver of this evolution. However, large portions of the uplands were disconnected from external drainage, routing surface flow to internally-drained lakes and wetlands. This landscape motivates the question of how drainage networks can expand when slopes are low and surface water is not necessarily routed toward channels.

We build a numerical model of post-glacial drainage network expansion based on the LandLab platform to explore the sensitivity of rates and patterns of fluvial network growth to flow routing over the land surface and via groundwater flow. Streams erode as a function of discharge and channel slope. We consider cases where surface flow is always forced out of closed depressions, where routing varies as a function of time or depression depth, and when water in closed depressions is lost to evaporation. In addition, we develop an idealized model of groundwater flow in which the groundwater divide is not dictated by surface topography, but instead is imposed to direct groundwater toward river valleys. Routing of surface and subsurface water is a dominant control on rates of channel growth: growth slows by orders of magnitude when water is not routed across subtle topographic divides. We consider a range of scenarios to identify changes in morphology that reflect increasing groundwater contributions. Specifically, groundwater contributions to stream discharge which are focused at a specific depth (stratigraphic layer) produce channel networks with profound changes in channel slope and degree of network branching associated with the depth of groundwater seeps. Observed rates of channel network expansion are more consistent with modeled evolution that includes routing of flow via surface or subsurface paths.