Baltimore Ecosystem Study Institute of Ecosystem Studies

2015 BES Annual Meeting Presentation and Poster Abstracts

Quantifying spatial patterns of channel geometry in an urban drainage network
Cole, Joshua
Co-Authors: Joshua Cole, Erin Stapleton, Andrew Miller, Claire Welty

Abstract: Quantifying spatial patterns of channel geometry in an urban drainage network Joshua N. Cole1, Erin Stapleton1,2, Andrew J. Miller1,3, and Claire Welty1,2 1Center for Urban Environmental Research and Education, University of Maryland Baltimore County, Baltimore, MD 21250 2Department of Chemical, Biochemical, and Environmental Engineering, University of Maryland Baltimore County, Baltimore, MD 21250 3Department of Geography and Environmental Systems, University of Maryland Baltimore County, Baltimore, MD 21250 Channel incision frequently occurs in urban watersheds. Incision is caused by high stormwater velocities and volumetric flow rates resulting from increased impervious cover and efficient storm drain networks shunting storm flow to urban stream channels. Channel incision has been shown to reduce groundwater levels in riparian zones, alter soil moisture characteristics, and reduce denitrification rates. Quantification of channel incision is often required for hydrologic, geomorphic, and biogeochemical assessments of urban stream systems, but standard techniques for such analyses are not currently available. As a result of a need to quantify incision in selected BES LTER watersheds, we have developed both automated and manual methods to select bank and thalweg points. These are used to characterize stream reaches and the adjacent valley floor and map the spatial pattern of channel incision within a watershed. Using GIS techniques, transects are created perpendicular to the channel at a high longitudinal frequency (e.g., 10m apart). These are used to extract elevation values from a LiDAR derived DEM. Approximate drainage area and distance upstream from the watershed outlet are also calculated and stored for each transect. Elevation transects are verified both visually and numerically. The method allows us to identify the edge of bank on both sides of the channel, identify thalweg location, and calculate parameters describing channel width, depth, and cross-section area. The results are saved in both graphical and tabular formats. Because there are multiple sources of uncertainty that may confound the automated selection of bankfull stage and bank edge purely from cross-section coordinates, a manual method for identifying these points is provided. This allows a user to exercise professional judgement in cases where the automated method is unable to find a satisfactory choice. Results can be analyzed for individual transects, averaged along a reach, or processed to identify longitudinal trends characterizing the locations and extent of channel incision. We applied these techniques to Horsehead Branch and Dead Run using analysis transects spaced 10 m apart. Horsehead Branch is a relatively undisturbed watershed characterized by 20% impervious surface area, whereas Dead Run is highly urbanized, with 53% impervious surface area. The median bank incision of HHB was found to be ~0.70 m, whereas Dead Run exhibited zones of highly variable incision, ranging from a median of about 0.62 m for least incised reaches to over 2 m at greatly incised reaches. Longitudinal incision data were found to exhibit distinctive periodic behavior for both streams. Geostatistical analysis of longitudinal incision data revealed correlation scales of about 55 m for HHB and 20 m for Dead Run, indicating that the spatial persistence of longitudinal incision patterns were much greater for HHB than Dead Run.