Search for: All records

Accelerated floodplain sedimentation related to agricultural development of uplands has produced postsettlement alluvium (PSA) along rivers throughout the upper Midwest, U.S.A. Landscape characteristics, surficial sediments, and soils in the region vary geographically in relation to differences in geologic history, yet the extent to which this geographic variability influences PSA accumulation remains unexplored. This study uses existing data to assess how non-dimensional PSA thickness varies with landscape characteristics, surficial sediments, soils and climate. Geographic variability is associated with three subregions: 1) areas glaciated during the Late Wisconsin Episode (LWE), 2) areas glaciated during Pre-Illinois and Illinois Episodes (PI&IE), and 3) the Paleozoic Plateau (PP), an area where evidence of Quaternary glaciation is highly localized and does not influence geomorphic characteristics of the landscape. These subregions differ significantly in average geomorphic characteristics, including mean watershed slope (WS), mean local relief (LR), fraction of non-contributing area (NCA), pre-settlement drainage density (DD), and mean normalized river steepness (KSN). Native vegetation type also differs systematically between the subregions, creating significant differences in the frequency of alfisols (Alfi) and molisols (Mol). Thickness of last glacial loess (Loess) also varies across the region, although not systematically between the subregions identified. Non-dimensional PSA thickness differs significantly among the subregions, increasing systematically with landscape age, reflecting faster upland erosion rates and stronger connectivity of uplands to river corridors in older landscapes relative to more recently glaciated landscapes. Nondimensional PSA thickness is significantly positively correlated with LR, KSN, WS, Loess, Alfi, and Mol and significantly negatively correlated with NCA. Non-visibly distinct PSA is present in some LWE watersheds characterized by significantly lower KSN and WS relative to other LWE watersheds in which PSA is visibly distinct. PSA thickness and visibility reflect catchment-wide landscape characteristics and watershed-scale river steepness, which emphasize the importance of geographic setting, geological history, and landscape geomorphic characteristics for understanding historical river sediment dynamics. Spatial variability in PSA thickness also serves as an indicator of river system sensitivity to land-use change, providing insight into the relative impact of humans on rivers within different geographic settings. 

Landscape dissection by rivers is a common qualitative measure of surface maturity. Quantitative studies of fluvial development over time indicate that drainage development increases non-linearly and is influenced by lithology, however, these studies typically take place over short timescales (10s of years), cover small areas, and focus on steep landscapes. In this work we use the Central Lowlands physiographic province (CL) as a natural laboratory in which we investigate rates and controls on drainage development in a post-glacial lowland landscape. Portions of the CL have been glaciated repeatedly in the Quaternary, and its topography is dominated by a patchwork of glacial landforms that have been developing drainage for 10 thousand to more than 500 thousand years. We modify the National Hydrography Dataset to estimate pre-agriculture drainage density developed over different amounts of time to reveal rates of drainage development in the CL. We find that drainage density in the CL increases non-linearly, increasing rapidly following glaciation before slowly approaching a maximum value. Much of the development is accomplished by 50 ka, well within a typical interglacial period. The apparent maximum value, ~1.5 km/km², is comparable to the median drainage density measured in regions in the CL that have not experienced Quaternary glaciation. Our study shows that this value is likely influenced by soil sand content and regional precipitation levels. We note that while drainage density increases to an apparent maximum within an interglacial, the fluvial network is unlikely to adjust to post-glacial base level conditions within that same length of time. Our results are most consistent with a model of drainage expansion driven by the connection of closed depressions, or ‘non-contributing area’ (NCA), the portion of a watershed that does not drain to a river. We find that NCA decreases in tandem with increasing drainage density, which implies that NCA could be a measure of landscape integration that is at least as sensitive as drainage density. 

In the low-relief post-glacial landscapes of the Central Lowlands of the United States, fluvial networks formed and expanded following deglaciation despite the low slopes and large fraction of the land surface occupied by closed depressions. Low relief topography allows for subtle surface water divides and increases the likelihood that groundwater divides do not coincide with surface water divides. We investigate how groundwater transfer across subtle surface water divides facilitates channel network expansion using a numerical model built on the Landlab platform. Our model simulates surface and subsurface water routing and fluvial erosion. We consider two end-member scenarios for surface water routing, one in which surface water in closed depressions is forced to connect to basin outlets (routing) and one in which surface water in closed depressions is lost to evapotranspiration (no routing). Groundwater is modeled as fully saturated flow within a confined aquifer. Groundwater emerges as surface water where the landscape has eroded to a specified depth. We held the total water flux constant and varied the fraction of water introduced as groundwater versus precipitation. Channel growth is significantly faster in routing cases than no-routing cases given identical groundwater fractions. In both routing and no-routing cases, channel expansion is fastest when ~30% of the total water enters the system as groundwater. Groundwater contributions also produce distinctive morphology including steepened channel profiles below groundwater seeps. Groundwater head gradients evolve with topography and groundwater-fed channels can grow more quickly than channels with larger surface water catchments. We conclude that rates of channel network growth in low-relief post-glacial areas are sensitive to groundwater contributions. More broadly, our findings suggest that landscape evolution models may benefit from more detailed representation of hydrologic processes.