Connecticut River-Barrett Brook Cove ReportFinal-2021 sandbarBarrett Brook Cove
Sand Bar Formation
Robert M Newton
Professor Emeritus, Department of Geosciences
Smith College
February, 2021
Figure 1. Air photo (5/6/2019) showing the location of Barrett Brook Cove on the Connecticut River. This is the
location of the Northampton Community Rowing docks and Barrett Brook enters the cove at the location of
the former Hampshire and Hampden barge canal.
Introduction
This report examines the recent formation of a sand bar at the mouth of Barrett Brook Cove on the
Connecticut River (Figure 1). Barrett Brook Cove is the location where the former Hampshire and
Hampden barge canal terminated in the Connecticut River. Construction of the canal was
completed in 1835 but it had a very short life, closing in 1847. However, with regard to this study, it
appears that at no time since 1847 has sand bar formation blocked the connection between the
Connecticut River and Barrett Brook. So why has a bar recently formed at this location?
Methods
River currents and bottom topography were mapped on August 6, 2019 using a Teledyne RD
Instruments RiverRay Acoustic Doppler Current Profiler (ADCP) equipped with a Hemisphere
A325 GNSS Smart Antenna. The survey was done with assistance from the Northampton Fire
Department who provided a small boat and able crew (Figure 2). As part of the survey, the
Figure 2. Northampton Fire Department
personnel helped with the survey. Here the
RiverRay is being operated by Matt Lemberg
while Steven Hall piloted the boat.
discharge of the Connecticut River was measured along
a transect at the Barrett Brook Cove. This measurement
assumed no moving bed as bottom tracking was used to
determine positions along the transect. In addition to
the discharge transect, a series of depth transects were
run across the river through the axis of the upstream
meander (Figure 3). The vertical beam of the ADCP
was used to determine water depth while GPS positions
were determined by a Hemisphere A325 GNSS Smart
Antenna directly connected to the ADCP. Depth
measurements were converted to bottom elevations by
subtracting the depth measurement from the elevation
of the water surface of the river. This elevation dataset
was merged with the 2015 QL1 LiDAR based DEM of
the local area obtained from MassGIS. A 1.45cm x 1.45cm raster was created from the point data
using a kriging technique and contours were created using the raster data. Because the RiverRay
collects very closely spaced depth measurements some had to be discarded prior to merging in order
to prevent transect line bias.
Figure 3. Data points collected on August 6, 2019 using a Teledyne RD Instruments RiverRay ADCP equipped
with a Hemisphere A325 GNSS Smart Antenna.
Results
• Discharge measured at 11:30am on August 6, 2019 was approximately 2000cfs.
• The lowest elevation in the channel bottom (68.6ft) occurs just opposite the Barrett Brook
Cove in a secondary depression in the river bottom (Figure 4). This is likely due to scour
associated with a large eddy that forms in that location.
Figure 4. Bottom elevations determined from the vertical beam of the RiverRay ADCP. Contours are elevations
measured in feet above mean sea level. Since the water surface elevation is usually around 102ft, water
depth = 102ft – bottom elevation.
• Even at these August low flow conditions, there is evidence for a large eddy in the Barrett
Brook Cove area (Figures 5 and 6).
Figure 5. Current direction and magnitude vectors collected along bathymetry transects. Note the eddy feature
near Northampton Community Rowing dock.
Figure 6. This cross section is oriented such that the left side is the left bank of the river and is located at about the
Northampton Rowing Club docks. The black line represents the bottom of the river and the arrows
represent the secondary velocity vector scaled to 33cm/sec. The colors represent primary downstream
velocity. The figure is based on over 200 ensembles with 12-23 bins in each ensemble.
The cross section in Figure 6 was created using the USGS Velocity Mapping Toolbox (VMT)
software (Pearson, et al., 2013) which uses data input from an ADCP to map the velocity field. In
this case, the colors represent the downstream component of velocity using the Rozovskii
definition (Rozovskii, 1957). This method rotates each measurement ensemble to zero the net
lateral flow over the entire ensemble. An ensemble is a vertical stack of velocity and depth
measurements (bins) taken by the ADCP. The RiverRay is capable of measuring up to 25 bins
simultaneously. This figure reveals an eddy on the right bank of the river in the area of the
Northampton Community Rowing Club docks.
• Sandbar formation at Barrett Brook Cove appears to be associated with an upstream
current associated with a large eddy that forms at that location. The evidence for this can be
seen in both the 2015 LiDAR based DEM (Figure 7) and in historical air photos
(Appendix).
Figure 7. LiDAR based DEM of the Barrett Brook Cove area with a 45˚ azimuth hillshade. Lighter colors (yellow and
browns represent higher elevations while the blue is the lowest elevation (Connecticut River). The old canal
can be clearly seen as the straight valley of Barrett Brook. The northern portion of the A-B cross section line
shows an accreting sand bar created by currents moving from right to left (upstream). The dashed yellow line
shows the original erosional scarp cut by the river.
The geomorphology on the east side of Barrett Brook Cove shows infilling of the cove from
sediment entering the system from the east. The original erosional scarp cut by the river is shown
by the yellow dashed line. Sediment influx from the east has created a bar that lies north of the
scarp, leaving a small remnant of the cove that is now isolated as a small wetland that is slightly
higher than the level of the cove (Figure 8). The “old bar” labeled on the topographic profile
(Figure 8) is vegetated and is separate from the most recently formed bar that is shown on the
5/6/2019 air photo (Figure 1). The north slope from the top of the bar to the Connecticut River is
not one continuous slope but is made up of at least 2 different slopes (Figure 8) suggesting multiple
bar forming events. An air photo taken on 5/6/19 shows recent bar formation (Figure 9). Figure
10 superimposes this bar onto the hillshade map from the 2015 LiDAR based DEM.
Figure 8. Topographic profile along the A-B line derived from the LiDAR based DEM of the Barrett Brook Cove
area. The Old bar was formed by sand migrating from east to west (upstream) across the area. The multiple
slopes on the Connecticut River side of the Old bar suggest multiple bar forming events.
Figure 9. Air photo of Barrett Brook Cove taken on 5/6/19 showing the sand bar growing from the
eastern side of the cove.
Figure 10. Hillshade of 2015 LiDAR based DEM with the bar shown on the 5/6/2019 air photo added.
• Underlying surficial geologic materials (Figure 11 and 12) are influencing the path of the
Connecticut River at this location. Normally the channel of a meandering river migrates
downstream through time as erosion on the cut bank on the outside of the meander occurs
just downstream of the meander axis. However, the sharp meander at the northern edge of
Northampton is “hung up” on till and perhaps bedrock that lies on the outside of the bend
(Figure 13). Meanwhile, the upstream meander continues to migrate (shown by blue arrow)
resulting in a “tight” downstream meander.
Figure 13. Sketch map of the Connecticut River showing how the Hatfield meander has migrated downstream
(blue arrow) while the Northampton meander is “hung up” on glacial till and bedrock (red unit).
Orange star marks location of Barrett Brook Cove.
N
Hatfield
Hadley
Figure 11. Portion of “Surficial Materials Map of the Easthampton Quadrangle, Massachusetts” compiled by Janet R.
Stone and Mary L. DiGiacomo-Cohen, 2019, USGS. The star shows the location of Barrett Brook Cove.
Figure 12. View of the right bank of the river just upstream from the Northampton Community Rowing dock. The
boulders along the shore indicate the presence of glacial till at this location.
Floodplain Alluvium
Lake
Sediments
Thin
Till
River
Terrace
Discussion
• The bathymetry of the Connecticut River in the vicinity of Barrett Brook Cove is unusual in
that the deepest part of the channel is opposite the cove rather than just downstream of the
apex of the meander as is normally the case. This is likely due a complex flow pattern that is
created by the sharp meander coupled with downstream flow around Elwell Island. It is
likely that a large eddy forms just off the cove that induces an upstream current near the
right bank of the river. The recent bar formation on the right bank is associated with this
eddy current.
• Analysis of historical air photos (Appendix) show periods of bar formation dating back to at
least 1995. Bar formation at the cove seems to be contemporaneous with deposition of sand
at the upstream end of Elwell Island. While deposition could be driven by small changes in
the channel of the Connecticut River, it could also be controlled by the amount of bedload
sediment being transported through this part of the Connecticut River system at any given
period of time. Channel erosion and deposition is not a continuous phenomenon but is
dynamic and constantly changing with changes in the river flow (discharge) and sediment
load.
• We might speculate that the cause of the recent bar deposition at Barrett Brook Cove is an
increase in bedload sediment transport associated with sand dumped into the Connecticut
River from the Deerfield River during Tropical Storm Irene in August of 2011. This event
was the storm of record on the Deerfield River and transported huge amounts of sediment
into the Connecticut River at Montague. The Connecticut River discharge was not
particularly high during this event, but the sediment load increased dramatically and a large
plume of suspended sediment was discharged into Long Island Sound. Little is known
about what happened to the sand sized bedload material. Since there are no dams to capture
the sediment between Montague and Northampton, it is likely that it has all been transported
downstream. The cove is located about 35km downstream from the Deerfield River so it
must have taken some time for the sand to move downstream. How much time is
unknown, but it is possible that it has taken several years and thus the appearance of the bar
at Barrett Brook Cove may reflect the arrival of the pulse of sediment associated with this
event.
Recommendations for Future Research
• There needs to be a hydrologic study of the pattern of flow in the Connecticut River in the
area of Barrett Brook Cove. This study should be done under a range of different flow
conditions. Not just at summer low flow as done during this study.
• The shape and volume of the newly formed bar should be monitored through time. This is
best done by drone based aerial photo analysis using “structure from motion” techniques. It
was hoped that we could complete an initial survey as part of this study, however COVID-
19 prevented that from happening. This survey should be done during a summer low flow
period when the bar has maximum exposure.
• A field study is needed to examine the surficial geology in more detail than is available in the
published surficial geologic map. The nature and extent of the material that is “holding up”
meander migration should be established. A seismic refraction line could be run to establish
depth to bedrock in the area. It is possible that the bedrock is exposed in the bed of the
river. Also the nature of the till should be established as there are two types of till that occur
in the Northampton area. The Lower Till is much more resistant to erosion than the Upper
Till and its presence could explain the channel morphology in this area.
References Cited
Parsons, D.R., Jackson, P.R., Czuba, J.A., Engle, F.L., Rhoads, B.L., Oberg, K.A., Best, J.L.,
Mueller, D.S., Johnson, K.K., and J.D. Riley, 2013, Velocity Mapping Toolbox
(VMT): a processing and visualization suite for moving vessel ADCP
measurements, Earth Surface Processes Landforms, v. 38, p. 1244-1260.
Rozovskii, I.L., 1957, Flow of water in bends of open channels, Academy of Sciences of the
Ukraine.
Stone, Janet R. and Mary L DiGiacomo-Cohen, 2019, Surficial Materials Map of the
Easthampton Quadrangle, Massachusetts USGS Geologic Quadrangle Report
Appendix of Air Photos (Google Earth