NATURAL CHANNEL DYNAMICS IN A GRAVEL BED STREAM IN EASTERN PA AS REVEALED BY PRE-DAM REMOVAL BASELINE MONITORING

We have been monitoring three segments of the Bushkill Creek in eastern Pennsylvania, in anticipation of the removal of three low-head dams (1.5 – 2.4 m high) since 2017. The Bushkill Creek watershed has an area of 207 square kilometers and flows south from the drainage divide formed by Kittatinny Ridge to its confluence with the Delaware River at Easton, PA. The upper half of the watershed is underlain primarily by the Martinsburg slate whereas the lower half of the watershed is underlain by limestone and dolostone. Although the stream flows over bedrock in a few sections, the majority of the channel bed in the lower section of the watershed consists of a mix of sandstone and carbonate gravel with the sandstone derived from a graywacke sandstone-rich member of the Martinsburg Formation and glacial till dating to MIS 6 (Illinoian). Standard shear stress calculations and Shields criterion for gravel mobilization indicates the stream typically experiences one to several discharge events per year capable of mobilizing the cobble-bed reaches in the lower watershed. Discharge data from a USGS gauging station in the watershed indicates that as many as seven bed mobilizing events occurred during the past two years and observations at specific locations reveals evidence of complete bed mobilization in some reaches.

We have been collecting data surrounding each of three low-head run-of-river dams located along the lowest 3.6 km of the stream. At each dam we have surveyed multiple cross-sections upstream and downstream of each dam, and measured bed material grain size at each cross section. Here we present data from 25 cross sections measured during the summers of 2019 and 2021. As might be expected, channel change was fairly minimal, however, measurable deposition occurred in the pool area of dam 1 (+0.258m). Average cross section-wide channel change for all 25 cross sections was +0.087m. Maximum cross section erosion was –0.14m and maximum deposition was +0.440m. Seven cross sections experienced slight erosion (average –0.068m) and 18 cross sections experienced net deposition (average +0.148m). These surveys provide data on natural variability in the channel bed associated with normal high flows in an undisturbed channel that will serve as the baseline data for comparison when the channel responds to dam removal in the near future.