U.S. Geological Survey Open-File Report 2011–1003
Combined Multibeam and LIDAR Bathymetry Data from Eastern Long Island Sound and Westernmost Block Island Sound – A Regional Perspective
The study area extends across eastern Long Island Sound and into westernmost Block Island Sound (figs. 1, 11). In general, shallow depths occur along its northern and southern edges and to the west, where thick marine deltaic deposits are found (Lewis and DiGiacomo-Cohen, 2000). The shallowest areas, which are essentially at sea level according to the LIDAR data, occur along shorelines of southeastern Connecticut (such as, Niantic Bay and New London Harbor) and Fishers Island, N.Y. Shallow areas also occur on top of isolated bathymetric highs, such as erosional outliers and shoals (for example, Six Mile Reef). The Sound narrows toward the eastern part of the study area, constricting the tidal flow and causing stronger bottom currents (Signell and others, 2000; White and White, 2009). These currents have scoured the sea floor and erosion has increased the water depths. The greatest water depths surveyed within the study area exceed 117 m and lie in scour depressions adjacent to the Race, and at the northern entrance to Plum Gut, the passage between Orient Point and Plum Island (fig. 11). The 4-m resolution of the DTM allows us to identify specific morphological features that indicate sea-floor geology, sedimentary processes, and anthropogenic impacts. Examples of features that can be identified in this dataset are reviewed in the following sub-sections.
Bedrock Outcrops and Glacial Moraines
Rocky areas are concentrated in the eastern part of the study area and along its shallow southern edge. Extensive tidal- and storm-driven currents over the past 15,000 years have removed or winnowed much of the glacial and younger sediment that once filled eastern Long Island Sound (Lewis and Needell, 1987; Lewis and DiGiacomo-Cohen, 2000; Signell and others, 2000; Knebel and Poppe, 2000; Poppe and others, 2000) and exposed bedrock outcrops and produced lag deposits of boulders.
Ice-sculpted bedrock is exposed at several locations in the northeastern part of the study area (figs. 12, 13, 14). Characteristics that were used to delineate the bedrock include outcrops that influence the morphology of the sea floor and have an associated pattern of distinctive parallel ridges on the DTM. The distinctive ridges on the bedrock surface can be correlated with those of onshore formations. The northwest-southeast bedrock trends (fig. 13) are similar to glacially smoothed bedrock ridges that trend between N. 15° and N. 20° W. on land immediately west of the Thames River. These elongate ridges, which are similar in size, orientation, and morphology to their inferred offshore counterparts, parallel the interpreted direction of ice movement as shown by grooves and striations (Goldsmith, 1962). The second set of ridges that trend northeast-southwest (fig. 13) are similar to features onshore in southeastern Connecticut, which trend between N. 65° E. and N. 80° E., and parallel the strike of foliation and the orientation of some jointing of the Avalonian anticlinorium (Rogers, 1985; Poppe and others, 2006b). These ridges and their offshore counterparts were probably formed by glacial plucking and resemble strike ridges formed in dipping layered rock. Together, these lineations show that the effects of glaciation on the bedrock are evident on the sea floor of Long Island Sound.
The northern flank of the Harbor Hill-Roanoke Point-Fishers Island end moraine segment underlies and outcrops along the southeastern edge of the study area (fig. 1; Sirkin, 1982; Lewis and Needell, 1987; Poppe and others, 2002). Many small, individual, rounded bathymetric highs give the sea floor in this area a rough appearance and are interpreted to be boulders (fig. 15). Although the boulders average less than 5 m in width, many exceed 10 m and the largest exceeds 20 m across. These rocky areas are lag deposits remaining from the winnowed sediments of the moraine.
Other recessional end moraines visible in the DTM include a previously unrecognized moraine in western Block Island Sound (fig. 16) and the Clumps moraine that extends westward off Fishers Island. Small, isolated rocky areas, which generally trend northeast-southwest (such as those at lat. 41° 12.8'N., long. 72° 37.5'W.; lat. 41° 12.8'N., long. 72° 30.1'W.; lat. 41° 13.2'N., long. 72° 22.0'W.; lat. 41° 13.15'N., long. 72° 18.4'W.; and lat. 41° 13.88'N., long. 72° 14.45'W.), are also visible in the DTM. We interpret these outcrops to be exposed moraine deposits because they follow the trends of recessional moraines mapped onshore across southeastern Connecticut (Schafer and Hartshorn, 1965; Flint and Gebert, 1976; Goldsmith, 1982; Stone and others, 2005) and, therefore, probably represent their offshore extensions.
Tidal currents, enhanced by constricted flow in the eastern sound, are responsible for high-energy sedimentary environments dominated by processes associated with erosion and non-deposition and have produced large scour depressions adjacent to the Race and at the northern entrance to Plum Gut (fig. 17). Bottom video from in the scour depressions (Poppe and others, 2010b) suggests that gravel armors much of the bottom in these features. However, gravel transported by currents may also be an abrading agent (Poppe and others, 2006b). As the swirling turbulent water in these fast eddies suspends and rolls gravelly sediment around in the depressions, impacts cause abrasion of the walls and floor. In this manner, erosion continues to enlarge and maintain the scour depressions.
Scour provides evidence of on-going erosion by the strong, oscillating tidal currents. For example, swirling currents develop around bedrock outcrops and boulders that protrude through surficial sediments. Turbulence causes increased flow velocity sufficient to scour the sea floor downstream from obstructions and to produce coarse-sediment floored sharp-pointed depressions called obstacle marks, also known as comet marks (Werner and Newton, 1975). Good examples of these bedforms are present around bedrock outcrops off Black Point and adjacent to the scour depression at the entrance to Plum Gut (figs. 14, 17). Asymmetry of the scour around these boulders indicates net transport; net transport is toward the direction with the greatest scour (Reineck and Singh, 1980).
Tidal energy drops as Long Island Sound widens westward (Signell and others, 2000), sedimentary environments dominated by processes associated with erosion give way to environments dominated by coarse bedload transport (Knebel and Poppe, 2000), and the rocky sea floor of the eastern sound is replaced by sands and gravelly sands (Poppe and others, 2000). Fields of megaripples (bedforms with 0.2 to 1 m of vertical relief) and sand waves (bedforms with greater than 1 m of vertical relief) are more common and larger under these slightly lower energy environments (figs. 17, 18, and 19).
Most of the sand waves in the eastern sound exhibit transverse morphologies (that is, elongation perpendicular to the prevailing currents). Crest-to-trough amplitudes of the largest transverse sand waves exceed 20 m, but these bedforms exhibit a range of amplitudes and crestline lengths. Current ripples occur on the flanks of the megaripples, megaripples occur on the flanks of sand waves, and sand waves occur on the flanks of larger sand waves (fig. 18). Many transverse sand waves exceed 1 km in length, and several are longer than 2 km (Poppe and others, 2008b).
Wave-crest morphology is variable, and sand-wave fields with smoothly sinuous, undulating, bifurcating, and relatively straight crest forms occur. Borders of sand-wave fields can be rather abrupt and are typically fringed with megaripples; megaripples occur more commonly on the stoss slopes of sand waves near the centers of the fields. Bifurcation of the wave crests, where present, is most common near abrupt changes in the character or orientation of the fields.
The sand waves reflect the strength of the bottom currents and can be used to indicate directions of net sediment transport (Allen, 1968; Ludwick, 1972). Except around shoals and along the southern side of the Race, most of the transverse sand waves display pronounced asymmetry with slip faces that are oriented toward the west and southwest and stoss slopes oriented to the east and northeast. This asymmetry suggests a flow-transverse crest orientation and net sediment transport toward the west and southwest. Wavelengths of the transverse sand waves vary by wave amplitude, but average about 150 m, and wavelengths of the largest transverse sand waves exceed 400 m (Poppe and others, 2008b).
The presence of current ripples and megaripples on the sand waves and the steepness of the slip faces suggest that this transport is active and that the sand waves are propagating under the present hydraulic regime (Dalrymple and others, 1978; Reineck and Singh, 1980). The presence of megaripples with crests oblique to trends of the underlying sand waves is evidence for flow separation and increased turbulence, which can promote increased sediment flux from the development of secondary flows (Allen, 1968). Whether the difference in orientation of sand waves and megaripples translates to variance in sediment transport directions or propagation time scales is uncertain.
Although current ripples may reverse their orientation during semidiurnal tidal cycles because of the small volume of the ripples, no reversal in either the sand-wave or megaripple morphology is apparent in the multibeam data of adjacent lines collected at different tidal stages. Whether this tidal independence of large bedform orientation is due to the spatial distribution of residual currents or to an asymmetry of current velocities is also uncertain, but it does suggest that the larger bedforms are more stable and that they move over longer time scales.
Transverse sand waves that are relatively straight crested, display no bifurcation, and are symmetrical in profile occur within a few small locations within the study area. Megaripples are typically absent from the stoss slopes of these waves, and their ends are marked by distinct scour zones or moats. These sand waves, which closely resemble bedforms described in Great South Channel and on the Canadian part of Georges Bank (Valentine and others, 2002; Todd and Valentine, in press), are probably immobile.
Although present, crescent-shaped barchanoid sand waves are much less common than transverse morphologies (fig. 19). Individual barchanoid sand waves average less than 100 m wide, but the largest exceeds 465 m. These waves commonly align into elongate fields having smaller isolated waves concentrated in the eastern ends. Typically, wavelengths increase, the fields widen, and the barchanoid waves become larger and coalesce into complex forms and highly curved ridges westward within barchanoid wave fields. The observed widening of the barchanoid sand-wave fields is probably caused by the propagation of younger waves off the horns of pre-existent waves situated up-drift in the transport regime.
Relief of individual barchanoid waves and complex forms averages 1-2 m and 3-5 m, respectively. However, isolated scour depressions, which cut into the underlying deposits, commonly occur in the troughs within the convex areas of the larger waves. If trough-to-crest amplitudes are measured from these depressions, the waves appear much larger, and individual barchanoid waves have average amplitudes of 2-3 m, complex forms that have coalesced into ridges have average amplitudes of 10-15 m, and some giant waves exceed 20 m. The presence of "blow-out" depressions is evidence for intense scour that occurs on the down-current sides of the barchanoid sand waves. This scour, which may be related to roll vortexes, supplies a mechanism for the continued erosion of underlying glaciolacustrine sediments and creates another source of fine-grained sediment for the sound.
Although the coalesced barchanoid bedforms retain their arcuate slip faces and relatively gentle stoss slopes, the horns are shortened, widened, and rounded. In places, barchanoid waves grade laterally into transverse morphologies. Regardless of size and complexity, all of the barchanoid sand waves, with the exception of those near the Race (Poppe and others, 2007a), are concave westward, indicating net westward sediment transport because the horns of a wave migrate more rapidly than the wave's main body (McKee, 1966).
The distribution of sand-wave types within the study area is controlled at least to some extent by the availability of sand. Transverse wave morphologies dominate where the sediment supply is more abundant, such as those associated with shoals or a large eroding marine deltaic deposit (Lewis and Needell, 1987; Lewis and DiGiacomo-Cohen, 2000; Poppe and others, 2002; Poppe and others, 2007b). Conversely, barchanoid wave morphologies typically form in environments where the substrate is firm and the sediment supply is limited (Reineck and Singh, 1980). Barchanoid sand-wave fields in the eastern sound coincide with areas where the Holocene section is thin (Lewis and DiGiacomo-Cohen, 2000; Poppe and others, 2002).
Aside from the features described above, some of the sea floor in the eastern sound appears to be relatively flat and featureless in the DTM (fig. 17). In these areas strong tidal currents prevent the deposition of Holocene marine sediments and erode the finer fraction from older sediments, leaving exposed lag deposits of gravel and gravelly sediment to armor the sea floor. Larger sessile forms are absent on these gravel pavements, suggesting that the gravel is periodically mobilized.
Anthropogenic influences on the bathymetry, such as shipwrecks, disposal sites, and infrastructure, can be seen at many places within the study area. Shipwrecks, which are of interest to historians and the diving community, are common. Examples of prominent shipwrecks include the previously uncharted wreck of the Volund, a 239-foot Norwegian tramp steamer, that lies just southeast of the Race in approximately 30 m of water (Poppe and others, 2006b), and the Lake Hemlock, a converted WWI cargo ship, that lies south of the western tip of Six Mile Reef (Poppe and others, 2008b). Asymmetrical scour around both wrecks shows that the oscillating tidal currents are strong and that transport is to the west at both sites.
Two disposal sites are visible on the DTM in this report. The New London Dumping Ground lies 1.8 to 4.1 km south of the entrance to the harbor channel (fig. 20). The dredge spoils and sediments used to cap them, which in places rise 5 to 6 m above the surrounding sea floor, give the sea floor there a rough, hummocky appearance. The Cornfield Shoals Disposal Site (Wiley, 1996; Poppe and others, 2011) is located about 6.1 km south of Cornfield Point, Conn. Although this is a dispersive disposal site (that is, material dumped here is expected to be transported out of the area by bottom currents), diffuse spoil mounds and circular impact structures can still be identified in the DTM. Vertical relief of the largest spoil mound exceeds 1.1 m above the surrounding sea floor. Other anthropogenic features in the DTM include the dredged New London Harbor Channel, mooring anchors, prop-scour depressions, buried cables, and bridge footings (fig. 20).
Acquisition and Processing Artifacts
Some of the features visible in the DTM are artifacts of environmental conditions, time limitations, equipment capabilities, and methodology during acquisition and processing. Rougher sea states during acquisition can result in minor noise and loss of resolution in the data, but artifacts due to time limitations and equipment capabilities are more conspicuous. For example, gaps of no data (white polygons) owing to inadequate multibeam coverage during surveys H11999 and H12013 are present in the southwestern part of the study area and off the mouth of the Connecticut River (figs. 19, 21). In shallower areas, where obstructions on the sea floor can more readily represent hazards to navigation (for example, survey H12013), these gaps have typically been filled with sidescan-sonar data (National Oceanic and Atmospheric Administration, 2009b). Artifacts can also result when datasets acquired with different types of equipment are merged. For example, the LIDAR data off New London and Niantic, Conn. (surveys H11224 and H11225) abut multibeam data (surveys H11441 and H11442), but the multibeam data are consistently 25-centimeters (cm) deeper than the overlapping LIDAR. Because 25 cm was within the vertical error budget for the LIDAR system and as reported for the Reson 8101 and 8125 systems installed on the NOAA launches, and to keep the 0-m depth contour (sea surface as obtained by the LIDAR system) at the MLLW sea level, no attempt was made to adjust the LIDAR bathymetry to match that of the other surveys. The result is a noticeable offset between these datasets (fig. 20; Poppe and others, 2010a).
Processing artifacts are largely confined to the edges of survey areas that overlap other surveys. Inasmuch as no attempt was made to adjust the bathymetric data, where discrepancies are present, narrow intermittent bands of noisy data result (figs. 13, 17). Although minor in the original bathymetric data, all of the above described artifacts are made more conspicuous by the 5x vertical exaggeration applied during production of the GeoTIFFs provided in this report.