USGS -- SMIG --
Surface-water quality and flow Modeling Interest Group

Bottom Currents and Sediment Transport in Long Island Sound: A Modeling Study

by Richard P. Signell, Jeffrey H. List, and Amy S. Farris

  USGS, Center for Coastal and Marine Geology
  Woods Hole Field Center
  384 Woods Hole Road
  Woods Hole, MA 02543-1598
  Internet: rsignell@usgs.gov, jlist@usgs.gov, afarris@usgs.gov
  Phone: (508) 457-2229, (508) 457-2343, (508) 457-2288
  FAX: (508) 457-2309

Editor's note:
This article will appear in the September issue of the Journal of Coastal Research. The document available here is based on the final draft provided to the editors. Minor discrepancies between this document and the published version, therefore, may exist.

This version of the article has all of the figures converted to thumbnails with links to the larger images. A version with all of the figures inline is also available; the download time may be longer, but the inline figures may be more convenient for viewing and printing.

Citation:
Signell, R.P., List, J.H, and Farris, A.S., 2000. Bottom Currents and Sediment Transport in Long Island Sound: A Modeling Study. Journal of Coastal Research, 16(3), 551-566.

Table of Contents

Abstract

A high resolution (300-400 m grid spacing), process oriented modeling study was undertaken to elucidate the physical processes affecting the characteristics and distribution of sea-floor sedimentary environments in Long Island Sound. Simulations using idealized forcing and high-resolution bathymetry were performed using a three-dimensional circulation model ECOM (Blumberg and Mellor, 1987) and a stationary shallow water wave model HISWA (Holthuijsen et al., 1989). The relative contributions of tide-, density-, wind- and wave-driven bottom currents are assessed and related to observed characteristics of the sea-floor environments, and simple bedload sediment transport simulations are performed. The fine grid spacing allows features with scales of several kilometers to be resolved.

The simulations clearly show physical processes that affect the observed sea-floor characteristics at both regional and local scales. Simulations of near-bottom tidal currents reveal a strong gradient in the funnel-shaped eastern part of the Sound, which parallels an observed gradient in sedimentary environments from erosion or nondeposition, through bedload transport and sediment sorting, to fine-grained deposition. A simulation of estuarine flow driven by the along-axis gradient in salinity shows generally westward bottom currents of 2-4 cm/s that are locally enhanced to 6-8 cm/s along the axial depression of the Sound. Bottom wind-driven currents flow downwind along the shallow margins of the basin, but flow against the wind in the deeper regions. These bottom flows (in opposition to the wind) are strongest in the axial depression and add to the estuarine flow when winds are from the west. The combination of enhanced bottom currents due to both estuarine circulation and the prevailing westerly winds provide an explanation for the relatively coarse sediments found along parts of the axial depression. Climatological simulations of wave-driven bottom currents show that frequent high-energy events occur along the shallow margins of the Sound, explaining the occurrence of relatively coarse sediments in these regions. Bedload sediment transport calculations show that the estuarine circulation coupled with the oscillatory tidal currents results in a net westward transport of sand in much of the eastern Sound. Local departures from this regional westward trend occur around topographic and shoreline irregularities, and there is strong predicted convergence of bedload transport over most of the large, linear sand ridges in the eastern Sound, providing a mechanism which prevents their decay. The strong correlation between the near-bottom current intensity based on the model results and the sediment response, as indicated by the distribution of sedimentary environments, provides a framework for predicting the long-term effects of anthropogenic activities.

Introduction

Long Island Sound is a major east-coast estuary located adjacent to the most densely populated region of the United States. Because of the enormous surrounding population, the Sound has received anthropogenic wastes and contaminants from various sources (Wolfe et al., 1991). As part of its National Coastal and Marine Geology Program, the U.S. Geological Survey conducted a regional study designed to understand the processes that distribute sediments and related contaminants in the Sound (Knebel et al., 1999). Studies of the sea-floor sedimentary environments (Knebel and Poppe, 2000), sediment texture (Poppe et al., 2000) and contaminants (Buchholtz ten Brink et al., 2000; Mecray and Buchholtz ten Brink, 2000; Varekamp et al., 2000) have revealed basin-scale variations in sediment and pollutant characteristics as well as local variations with scales of several kilometers. Because bottom currents generate the shear stresses that resuspend and transport material from the sea-floor, knowledge of the bottom-current regime is crucial both in understanding the distribution of bottom sedimentary environments in the Sound and in predicting the long-term fate of wastes and contaminants which have been introduced there.

Previous work has described many of the major aspects of physical processes in Long Island Sound. Major reviews can be found in Riley (1956), in the Long Island Sound special issue of Advances in Geophysics (Bokuniewicz, 1980; Gordon, 1980; Bokuniewicz and Gordon, 1980a,b), and in the Long Island Sound Study final report (Schmalz et al., 1994). The purpose of this paper is to build on this previous work, using process-oriented numerical studies to: (1) elucidate the role that tide-, density-, wind- and wave-driven bottom currents play in the formation of sea-floor environments, and (2) examine regional and local bedload sediment transport pathways in the Sound. Signell et al. (1998) reported on the initial findings of the tide-, wind- and wave-driven components of this study.

Environmental Setting

Long Island Sound is located between Connecticut and Long Island, New York, on the east coast of the United States (Figure 1). It is approximately 150 km long, 30 km wide, and the average water depth is 24 m. A 30-60 m deep axial depression runs east-west through the central and western parts of the Sound, and water depths reach more than 100 m at the eastern entrance to the Sound. The Sound connects to the waters of Block Island Sound on its eastern end, through a constriction called "The Race" (northeast of PI in Figure 1). The Sound is connected to New York Harbor through the East River on its western end (just west of WP in Figure 1). The system is approximately in quarter-wave resonance with the semi-diurnal tide, resulting in a threefold increase in the average tidal range from 0.8 m on the eastern end to more than 2.2 m on the western end. Freshwater input is dominated by the Connecticut River (discharge is near SJ in Figure 1); its average freshwater input is 560 m3/s. The Housatonic and Thames Rivers (near BP and NL in Figure 1) add 130 m3/s of freshwater, and there is an estimated 10-60 m3/s input of freshwater to the Sound through the East River (Jay and Bowman, 1975). Winds are generally from the northwest during winter and from the southwest during summer, with the most energetic wind events typically occurring during the winter months.

fig. 1
Figure 1. Long Island Sound bathymetry. The average depth is 24 m. Locations of tide gauges used for model assessment are shown as crosses. See Table 1 for tide gauge station names and locations.

Density Structure

The density structure plays an important role in generating and modifying the bottom currents in the Sound. The vertical density structure varies seasonally with the annual surface heat flux and river inflow cycles (Riley, 1956; Schmalz et al., 1994). During winter, strong surface cooling prevents significant vertical stratification in temperature. Thermal stratification develops during April over most of the Sound, and by June, top-to-bottom differences of 3-5o C are common. The exception is near The Race, where vertical stratification is slight throughout the year due to strong tidal mixing (Bowman and Esaias, 1981). Overturning generally occurs in late August or early September, destroying the thermal stratification. Slight salinity stratification generally persists throughout the year, with typical top-to-bottom differences of 0.5-1 ppt. The regional salinity stratification is enhanced in the spring due to high freshwater discharge, and very strong localized stratification is associated with thin buoyant freshwater plumes (Garvine, 1974). Although there is considerable variability and complexity in the distribution of salinity in the Sound, a longitudinal salinity gradient of 5-6 ppt persists over the length of the Sound throughout the year (Kim and Bokuniewicz, 1991).

Tidal Currents

Tidal currents are the dominant currents throughout most of the Sound. The largest tidal current constituent is M2 (12.42 hour period), and the other semi-diurnal constituents S2 and N2 are each about 20% of M2 (Schmalz, personal communication, 1997). At perigean spring tides, therefore, the tidal currents can be 40% stronger and, during apogean neap tides, they can be 40% weaker. The diurnal current constituents K1 and O1 are quite weak, only about 5% of M2. Early measurements showed that, due to the large amount of water that must flow through the constricted opening at the Race, peak surface tidal currents are typically 120 cm/s and exceed 160 cm/s during perigean spring tides (Lelacheur and Sammons, 1932). The tidal current speeds decrease markedly westward with distance from the mouth as the width of the Sound expands, and tidal speeds of 20-30 cm/s are common over much of the Sound. As part of the EPA/NOS Long Island Sound Study (LISS), acoustic Doppler current profiler (ADCP) data was collected at 21 different locations (Earwaker, 1990) and subjected to harmonic current analysis. This analysis showed that internal tides could be present in the vertically stratified region of the Sound during summer, and act to enhance the tidal currents at the bottom. Due to the shallow water depths, strong tidal currents, bottom friction, and the complex geometry, the tides can generate significant residual currents and ebb/flood asymmetry. In Long Island Sound, as in most estuaries without substantial mudflats (Speer and Aubrey, 1985), the frictional decay of the tidal wave causes the flood tide to be shorter than the ebb tide (Redfield, 1980). The flood is about 15 min shorter than the ebb at New London (NL), and about 30 min shorter at Willets Point (WP). When averaged over any cross section of the Sound, observed flood tidal currents should be slightly stronger than the observed ebb currents. In addition to the ebb/flood asymmetry, Ianniello (1981) predicted there should be a 2-3 cm/s bottom flow toward The Race in the eastern Sound, based on idealized modeling of the tide-induced residual circulation. Observational evidence for both of these tidal-current effects have been difficult to establish from current meter measurements, however (Schmalz et al., 1994), perhaps due to the tidal asymmetries and tide-induced residual flow induced by local topographic features in the vicinity of the current-meter moorings.

Numerical models have been quite successful at describing the overall tidal characteristics. Kenefick (1985) showed that a 1-km resolution, depth-averaged, numerical model was capable of reproducing most of the phase and amplitude structure of the M2 tide. Schmalz et al. (1994) showed that a 2.2 km resolution three-dimensional model could also reproduce the phase and amplitude structure of both semidiurnal and diurnal constituents. These models were not of sufficient resolution, however, to describe the strong variations in tidal currents that occur at scales of a few km or less in the eastern Sound.

Density-Driven Circulation

The longitudinal salinity and associated density gradient that exists throughout the year drives an estuarine circulation, where water flows eastward near the surface and westward near the bottom (Riley, 1956; Wilson, 1976). Using data from 28 current meter moorings located 2 m above the bottom, Gordon and Pilbeam (1975) calculated an average westward bottom flow of about 5-10 cm/s. Superimposed on this long-term mean estuarine circulation are density-driven currents with strong spatial and temporal variability. Variations in river discharge and coupling between the wind and buoyancy inputs create complex density and velocity structure which makes it difficult to relate residual-current variability to external factors such as wind or river discharge (Gordon and Pilbeam, 1975; Schmalz et al., 1994). For example, Riley (1956) postulated the existence of a counterclockwise gyre in the western Sound and a clockwise gyre in the central Sound. Using extensive hydrographic measurements, ADCP data, and modeling from 1988-1990, Schmalz et al. (1994) showed that although gyral structures sometimes exist, they are not persistent and can be destroyed by moderate winds. River plumes generate intense current variability, but this variability is chiefly limited to the surface waters as the plumes are typically 2-4 m thick (Garvine, 1974).

Wind-Driven Circulation

In long, semi-enclosed basins, both local and remote winds drive currents. The local effect is caused by wind along the axis of the basin that sets up a surface elevation slope; circulation results from the balance of the along-axis wind, pressure gradient, and bottom friction (Csanady, 1973). This water flows downwind in the shallows and returns against the wind at depth. The second process is a remote effect, where the wind stress acting over the adjacent continental shelf acts to raise or lower regional sea level. These rising and falling water levels generate currents within the Sound that are sometimes correlated with the local wind. In bays with large surface area and constricted entrances, such as Chesapeake Bay, these remotely driven currents can produce large fluctuations in the vicinity of the bay entrance (Wang, 1979). Gordon and Pilbeam (1975), however, could not find any direct correlation between wind and bottom current in the Sound, and Bokuniewicz and Gordon (1980a) concluded that most storms (they note the exception of Hurricane Belle) do not generate significant residual currents in the lower half of the water column. Despite the lack of observations showing a coupling between the wind and bottom currents in the Sound, there is some evidence of wind influencing the bottom currents from previous modeling work. Schmalz (1993) decomposed the residual circulation for a one-month period (September 1988) and showed that the local eastward wind stress contributed significantly to the mean westward bottom circulation in the central Sound, whereas the contribution of the remote wind forcing was negligible.

Wave-Driven Currents

In their assessment of sediment transport and deposition processes in the Sound, Bokuniewicz and Gordon (1980a) used simple fetch and duration dependent wave theory, developed for reservoirs, to estimate the influence of surface waves on the sea-floor sediments. They determined that, due to the limited fetch and short duration of strong wind events, bottom wave-orbital velocities rarely exceed 10 cm/s in water depths greater than about 18 m. They concluded that is it unlikely that sea-floor sediments will be significantly affected by surface waves except around the shallow margins of the Sound and over shoals, where the water is less than 18 m deep. They supported these theoretical calculations with wave observations in the western Sound, where they observed a maximum wave-orbital velocity of 42 cm/s at 12.5 m water depth during sustained winds of 15 m/s along the axis of the Sound (the theory predicted 64 cm/s). Analysis of the seabed in the central Sound after Hurricane Gloria showed that 1-2 cm of muddy sediment had been resuspended and a 5-6 cm thick sand layer was deposited over mud offshore of New Haven Harbor (Rhoads, 1994). Even when storm waves do not cause sufficient bottom currents to resuspend material, they can enhance resuspension caused by tidal or other currents via wave-current interaction (Grant and Madsen, 1979).

Sedimentary Environments

Knebel and Poppe (2000) outline the general distribution of modern sea-floor environments in Long Island Sound. They identify four categories of environments based on an extensive regional collection of sidescan sonar and supplemental verification data. These categories include:
  1. erosion or nondeposition;
  2. coarse-grained bedload transport;
  3. sediment sorting and reworking; and
  4. fine-grained deposition.
In the funnel-shaped eastern part of the Sound, they find a westward progression of bottom environments ranging from erosion or nondeposition at the narrow eastern entrance to the Sound, through an extensive area of bedload transport and sediment sorting, to a region of fine-grained deposition. The broader central and western Sound, on the other hand, is comprised largely of depositional environments, except in local areas of topographic relief where there is a patchy distribution of various other environments. An extensive treatment of the bottom sedimentary environments in Long Island Sound may be found in Knebel et al. (1999) and Knebel and Poppe (2000). These analyses also indicate that winnowing of sediments occurs along the shallow margins and along some segments of the axial depression of the Sound.

Although previous studies have qualitatively described many of the relationships between physical processes and sea-floor characteristics in the Sound, we seek to answer several specific questions:

  1. Is there a well-defined boundary in tidal-current strength at which the transitions between environments occur?
  2. Is there a well-defined boundary in wave-current strength between environments?
  3. What processes prevent the filling of the axial depression with fine-grained material?
  4. How are the sand ridges in the eastern Sound maintained in the presence of strong tide- and wave-generated currents?
We use high-resolution (300-600 m) wave and circulation models to conduct simulations that address these questions.

Methods

Circulation Modeling

To address the bottom currents associated with tides and strong wind events, we configured a high-resolution model of Long Island Sound capable of representing topography at the 1-2 km scale. We used the Estuary Coastal and Ocean Model (ECOM) (Blumberg and Mellor, 1987) with 10 levels (each 10% of the water column) and 300 x 100 grid cells in a curvilinear domain (Figure 2). This resulted in a typical grid spacing of 200-400 m over most of the Sound.

fig. 2
Figure 2. The curvilinear model grid contains 300 x 100 cells, with typical cell sizes between 200 and 400 m. The grid shown is subsampled by a factor of four for clarity.

For tidal open-boundary conditions at the eastern end, we specified elevation with M2 tidal constituent data interpolated from the detailed ADCIRC finite-element tidal model of the East and Gulf Coast (Luettich and Westerink, 1995). The ADCIRC boundary conditions were further modified so that model predictions matched the observed tidal elevation and phase at Montauk (labeled MP in Figure 1). This required reducing the ADCIRC tidal amplitudes by 15% and increasing the phases by 9o. For the simulations with wind, we used a uniform wind stress, and forced the open boundary of the model with elevations obtained from steady wind runs with ADCIRC.

The bottom friction in numerical tidal models is usually treated as a "tuning" parameter, with the value adjusted until the best fit to the major elevation or velocity tidal constituents is achieved. Often a uniform value is used due to lack of spatial information concerning the nature of the seabed. In our study, however, we had access to high-resolution maps of the sedimentary environments, and attempted to obtain a best-fit bottom friction in each of the four main sedimentary environments through a parameter search. It turns out, however, that the amplitude and phase of the M2 elevation field is completely dominated by the friction in the erosional and bedload-transport environments in the eastern Sound, and is insensitive to the friction specified in the depositional environments in the central and western Sound. This is consistent with the finding by Bokuniewicz and Gordon (1980b) that the frictional dissipation in the eastern Sound is seven times greater than the dissipation in the central and eastern Sound. The optimal value of the bottom roughness length z0 for the erosion and bedload regimes was determined to be 1.0 cm, the same value used in the study by Schmalz et al. (1994). Since we could not determine the spatial variability in z0 from the elevation data, we used z0=1.0 cm throughout the Sound. The model was run for 5 tidal cycles to reach periodic equilibrium, with results saved every 10 lunar min over the last cycle. An internal time step of 74.52 s was used, with an external time step of 6.21 s. The coefficient in the Smagorinsky (1963) horizontal viscosity parameterization was set to 0.05.

For the tide and wind-driven simulations, the model was run with uniform density. Although vertical stratification exists for much of the summer and can modify both the tidal- and wind-driven response, the barotropic tide captures most of the tidal response, and the strongest wind events typically occur during the relatively unstratified winter months. For the simulations of estuarine circulation, the model was initialized with a longitudinal salinity variation of 6 ppt, consistent with the observations analyzed by Kim and Bokuniewicz (1991), and it was allowed to adjust over the 5 initial tidal cycles. These simulations provide a qualitative representation of the basin-scale gradients in the observed net westward bottom flow.

Wave Modeling

To better understand the resuspension potential throughout the Sound, we simulated the patterns of bottom orbital currents in the basin with the numerical wave-prediction model, HISWA (HIndcasting Shallow water WAves, Holthuijsen et al., 1989). HISWA computes steady-state wave heights on a rectangular grid over complex topography. It includes the simultaneous effects of wave generation by wind, wave propagation including shoaling and refraction, and wave dissipation through bottom friction and breaking. Incoming wave energy from outside the model domain was assumed to be zero due to the nearly enclosed nature of the Sound.

A square computational grid was constructed with dimensions 220 x 220 km. The grid spacing was 300 m in the direction with the wind and 600 m in the direction perpendicular to the wind. This grid was centered on Long Island Sound, allowing prediction of waves generated by wind from all points of the compass. We computed 144 HISWA simulations of the bottom wave-orbital velocity maximum, Ub, for winds of 2.5, 5.0, 7.5, 10.0, 12.5, 15.0, 17.5, 20.0 and 22.5 m/s, for each of 16 directions equally spaced around the compass.

In order to calculate a long-term climatology of bottom wave-orbital velocity (Ub) throughout the region, the set of 144 model simulations of bottom orbital velocity were weighted with the wind distribution over a 12-year period (Nov 1984 - Dec 1996) from the NOAA Ambrose Light meteorological station. From this calculation, one can predict the occurrence of orbital velocities exceeding a critical threshold or describe the distribution of wave-orbital velocities at a given location.

Bedload Transport Modeling

To investigate potential sand transport pathways and the maintenance of the sand ridges in the eastern Sound, we calculated bedload transport rates using the empirical formula presented by Gadd et al. (1978):

q = B (u100 - ucr)3

where q is the sand transport rate in g cm-1 s-1, B is a parameter that depends on grain size with units of g cm-4 s2, u100 is the flow speed at 100 cm above the bed, and ucr is a critical velocity threshold (also dependent on grain size) below which no sediment transport occurs. There are many bedload transport formulas available, and this is a particularly simple one. It serves to illustrate the combined impact of the tides and estuarine bottom flow that provides a persistent mechanism for moving sand in the Sound. We used values appropriate for medium-fine sand of 0.225 mm grain size, the same values used by Harris et al.. (1992) in their study of subtidal sand bank evolution in eastern Australia: ucr =17 cm/s, B = 5.8x10-5 g cm-4 s2.

Results

Tidally-Driven Bottom Currents

To verify that our simulations have the correct basic tidal structure in the Sound, we performed harmonic analysis on the model output and compared the dominant M2 characteristics at 9 different locations for elevation (Figure 1, Table 1) and 13 locations for velocity (Figure 3, Table 2). Because we are representing the barotropic tide in the tidal model, and because many of the velocity observations were obtained during summer when baroclinic tidal effects were evident, we compared the modeled M2 depth-averaged currents to the M2 depth-averaged currents obtained from the analysis of the NOS ADCP data. To compute the depth-averaged M2 major axis from the data, we weighted the analysis supplied by NOS (Schmalz, personal communication, 1997) for each depth level by the percentage of the water column it represented.

table 1
Table 1. Comparison of data and model M2 tidal elevation amplitude and phase in Long Island Sound.

fig. 3
Figure 3. Simulated depth-averaged M2 major axis tidal-current magnitude. Locations of NOS "RADS" stations are also shown, where velocity data for model assessment is available via ADCP measurements. See Table 2 for station information.

table 2
Table 2. Comparison of data and model M2 tidal current major axis amplitude in Long Island Sound.

The results for elevation show that the mean amplitude error is 1%. The modeled amplitude is within 4 cm of the data at all locations except New Haven, which is off by 8.9 cm, or about 8%. Kenefick (1985) and Schmalz et al. (1994) also noted New Haven as having anomalously large model-data error. They attributed this to the fact that the tide gauge is within a harbor, where it is possible there is a local influence not resolved by the models. The mean phase error for elevation is 2.2o, and at all locations the phase error is less than 7o (about 15 min).

The results for the M2 currents show that the model is within 10% of the observed speeds for 6 of the 8 stations west of The Race. The model does a good job at representing the depth-averaged M2 flows within the central and western Sound (Stations 4, 6, 7, 8 and 9) with a mean error of 2.3 cm/s, within 10% of the observed flow speeds. The modeled flows are slightly weaker than the observations with the exception of station 8. In the eastern Sound (stations 13, 20, 21 and 22), the tidal flows are stronger, and the geometry and tidal structure are more complex. At station 13 near The Race and at station 21 on the south side of Long Sand Shoal, the modeled current is about 20% too weak, whereas at stations 20 and 22 to the west and east of Long Sand Shoal, the tidal current is only about 6% too weak. East of The Race, the modeled speeds are within 5% of the data in the passage to the open ocean (stations 14 and 16) but are about 50% weaker than observed in the passage north of Block Island. This is most likely due to phase errors in the open boundary forcing that drive the model. The model forcing would need to be improved if used to assess currents in the vicinity of Block Island. It is curious that the depth-averaged M2 currents for all stations west of The Race (the region of interest) average about 10% lower than the observed values, because we know from the comparison of M2 elevation that the modeled error in the overall volume of water entering and leaving the Sound each tidal cycle is only 1%.

With the generally good performance of the depth-averaged currents established by the model-data comparison, the maximum bottom velocity (1 m above bottom) over the course of the tidal cycle was calculated as an indicator of the intensity of currents driven by typical tides. The results show strong bottom currents in excess of 50 cm/s in the constricted eastern end of the Sound, with the peak speed decreasing westward as the width of the Sound increases (Figure 4). In general, the eastern third of the Sound has bottom tidal speeds of 30-60 cm/s, the central third of the Sound has speeds of 20-30 cm/s, and the western third of the Sound has speeds less than 20 cm/s. Local enhancements of the bottom tidal currents exist near headlands and atop the cross-Sound shoal complexes in the western Sound; in places the currents exceed 30 cm/s. The typical spring/neap cycle can generate currents that are 20% stronger or weaker than the M2 speeds shown. During perigean spring tides, the bottom currents can be as much as 40% stronger.

fig. 4
Figure 4. Simulated maximum near-bottom tidal current speed (1 m above bottom) during periods of average tidal range. Spring tides are 20-40% stronger and neap tides are 20-40% weaker.

There is a clear correspondence between the weakening of the tidal currents westward of the Race and the westward progression of sedimentary environments in the eastern part of the Sound as outlined by Knebel and Poppe (2000). With the high-resolution information on both the sedimentary environments and the tidal currents, we can determine the distribution of bottom tidal currents within each environment. For the erosion and sediment sorting/reworking environments, there is a bimodal distribution of tidal currents, whereas for the bedload transport and depositional environments there is a unimodal distribution (Figure 5.) Also, there is a regular progression of the current strength among the four environments (see right-most in Figure 5). This progression corresponds to the regular change of sedimentary environments west of the Race. For the erosion environments, the part of distribution dominated by currents of 40-50 cm/s corresponds to the large area of erosion at the eastern entrance to the Sound, whereas the remaining part of the distribution with tidal currents less than 20 cm/s corresponds mostly to the shallow margins of the Sound, where wave processes play a dominant role (discussed below). For the bedload transport environment, the bulk of the distribution is described by 35-45 cm/s tidal currents, corresponding to the large region of coarse-grained transport just west of the large erosional area. Moving to the sediment sorting and reworking environment, the part of the distribution dominated by currents of 25-35 cm/s corresponds to a large band of sediment sorting located just to the west of the bedload transport region in the eastern Sound. The remainder of this distribution corresponds to areas dominated by local processes, such as on the flanks of shoal complexes, across areas of subtle elevation variations, and within the axial depression. Finally, tidal currents between 15-25 cm/s are typical of fine-grained depositional environments. Even though long-term deposition is dominant in these areas, it is evident that tidal currents can frequently be sufficient to resuspend the fine-grained material, especially when considering the 20-40% strengthening of the currents during spring tides. This inference is consistent with observations that show persistent benthic turbidity zones extending several meters or more above the bottom over much of the muddy areas in the central and western Sound (e.g. Rhoads, 1994).

fig. 5
Figure 5. Distribution of maximum near-bottom tidal current speed for each of the four major sedimentary environments described by Knebel and Poppe (2000).

The harmonics and residual currents generated when oscillatory tidal currents interact with variations in basin geometry can be important factors in net sediment transport. It is also important to document the tide-induced residual flow because it can confound attempts to measure or isolate residuals associated with other processes, such as the density-driven estuarine circulation. The most striking feature of the near-bed tide-induced residual flow is the 4-8 cm/s clockwise circulation around Long Sand Shoal in the eastern Sound (Figure 6). Other small-scale tide-induced circulations are seen in the vicinity of The Race and around coastal headlands. There is little evidence of a broad eastward tide-induced bottom flow in the eastern Sound, predicted by Ianniello (1981).

fig. 6
Figure 6. Simulated near-bottom (1 m above bottom) M2 tide-induced residual currents.

Wave-Driven Bottom Currents

In addition to tidal currents, the orbital currents associated with waves generated by local winds can be a significant mechanism of bottom sediment resuspension. An example of predicted Ub for winds of 15 m/s from the east-northeast (typical of a strong winter northeaster) is shown in Figure 7. Under these storm conditions, the significant wave height ranges from 1.5 to 2 m, with typical periods of 4-6 seconds. The bottom velocity ranges from less than 5 cm/s in water deeper than about 20 m to more than 20 cm/s in water shallower than about 10 m, generally found within a few kilometers of the coast. The wave velocity necessary to resuspend fine-grained muds is approximately 15 cm/s (Komar and Miller, 1975). Thus wave-induced bottom velocities during strong wind events could explain the winnowing of fine-grained sediments observed along the shallow margins of the Sound.

fig. 7
Figure 7. Simulated bottom wave orbital speed for steady east-northeast wind of 15 m/s.

Using the climatological approach discussed earlier, we can compute the percentage of time over the 12 years of wind observations that the wave orbital velocity is predicted to exceed a certain threshold value. An example is given in Figure 8 for a threshold of 15 cm/s. Similar to the northeasterly storm example, the percentage of time that Ub exceeds 15 cm/s is greatest in a thin strip around the periphery of the Sound. This threshold value is exceeded less than 0.001 percent of the time (less than 8 hours/yr) in water depths greater than about 20 m. This is roughly consistent with estimates of wave influence by Bokuniewicz and Gordon (1980a).

fig. 8
Figure 8. Percentage of time that the root mean square bottom wave orbital velocities exceed 15 cm/s, based of 12 years of wind data from the NOAA Ambrose Light meteorological station.

As with the tidal currents, we can use the high-resolution description of the sea-floor environments and wave model results to reveal the aspects of the wave climate that characterizes each environment (Figure 9). It is clear that the largest wave influence is in the erosion or non-deposition environments, and in particular, in those environments that are found along the shallow margins of the Sound.

fig. 9
Figure 9. Distribution of percent occurrence of 15 cm/s bottom wave-orbital speed for each of the four major sedimentary environments described by Knebel and Poppe (2000).

Wind-Driven Bottom Currents

In addition to forcing surface waves, strong wind events in the Sound generate bottom currents that may influence the distribution of sedimentary environments. Similar to the steady wind response in a long lake (Csanady, 1973), the currents in Long Island Sound respond most efficiently to the along-axis wind component. Wind blowing along the axis of the Sound rapidly sets up a local tilt of the sea surface opposing the wind; circulation is generally downwind in the shallows and against the wind in the deeper parts of the basin. Figure 10 shows the near-bottom current response to a 10 m/s westerly wind blowing along the axis of the Sound. Subtracting the tide-induced residual circulation has isolated the wind-driven residual circulation. Along the shallow margins there are strong bottom currents in the downwind direction. The deep return flow, opposite the wind, occurs mostly in the axial depression, where the flow speed is 6-8 cm/s. On either side of the axial depression, flow speeds weaken dramatically. In most of the central and eastern Sound, there is only a weak response. This is consistent with theory of Hunter and Hearn (1987), who showed that the wind-driven transport in long, shallow lakes is determined largely by the distribution of depths comprising the channel cross-section. They found weak transport when the variations in depth were small compared to the mean depth (e.g. when the cross-section is nearly constant depth). Shallow depths and large variations in depth, such as occur in the axial depression, lead to large transports and stronger bottom currents.

fig. 10
Figure 10. Simulated near-bottom currents (1 m above bottom) during a moderate westerly wind event (10 m/s). The tide-induced residual currents have been subtracted to isolate the residual currents due to wind.

In the axial depression, winds from the west drive a westward current that adds to the westward near-bottom estuarine inflow. Thus, westerly winds (the predominant wind direction) act to further enhance flows in the axial depression relative to the flows in the surrounding shallower regions. In contrast, storm winds from the east drive an eastward-directed bottom current that opposes the estuarine flow and, therefore, decreases the magnitudes of the currents in the depression. From analysis of the Ambrose Light wind data, westerly low-frequency wind events having wind speeds of at least 10 m/s occur about 10-20 times a year chiefly during the winter months. Although Bokuniewicz and Gordon (1980b) could not find a direct coupling between wind events and bottom currents in the shallower regions of the Sound, the bottom-current fluctuations in the axial depression are strongly correlated with the along axis wind. For example, at the NOS RADS station 7, located in the axial depression, 62% of the variance in the along axis subtidal flow was explained (during the summer 1990 deployment) by the along-axis subtidal wind component (Figure 11). The transfer function between wind and current was approximately 1% at this level (e.g., a 10 m/s wind drives a 10 cm/s current), in rough agreement with our model simulation. Also apparent in Figure 11 is a strong westward mean current of 11.3 cm/s. This strong mean flow cannot, however, be explained by the mean eastward wind component, which is just under 1 m/s for this time period. It is most likely the result of the enhanced density driven circulation that occurs in the axial depression.

fig. 11
Figure 11. Comparison of subtidal, along-Sound wind and current 4 m above bottom at NOS RADS Station 7 located in the center of the axial depression (see Figure 3 for location). The majority of the subtidal current variance (62%) is explained by along-Sound wind variations, with an approximate transfer function of 1 cm/s of current for 1 m/s of wind.

Density-Driven Estuarine Circulation

Observations and theoretical studies have established the presence of westward bottom flow in the lower layer of Long Island Sound due to the longitudinal salinity gradient. Just as the pressure gradient established by wind blowing along the Sound drives enhanced flow in the axial depression, the estuarine circulation in our simulation indicates strongly enhanced flows due to the along-Sound baroclinic pressure gradient (Figure 12). The strongest response tends to follow the thalweg of the Sound, with 2-4 cm/s flow running through the eastern and much of the central Sound, but increasing to 6-8 cm/s in the axial depression. Since this density-driven circulation persists throughout the year (albeit with much variability), it plays a crucial role in the long-term net transport of sediment. The local enhancement of flows in the axial depression by the estuarine circulation, augmented by westerly wind events provides an explanation for the relatively coarse material found along parts of the axial depression.

fig. 12
Figure 12. Simulated near-bottom density-driven currents (1 m above bottom). The tide-induced residual currents have been subtracted to isolate the residual currents due to the along-Sound density gradient.

Bedload Transport of Sand

Our high resolution, three-dimensional, nonlinear tidal model gives us the opportunity to predict sand transport pathways on the scale of 1 km. The persistent currents in the Sound are the tides and the estuarine circulation. Using tidal currents driven by M2 forcing and the estuarine circulation, and the method of Gadd et al. (1978), we compute the net bedload transport of medium sand averaged over a tidal cycle (Figure 13). As expected, the eastern Sound shows by far the greatest potential for sand transport sand, and since tides are significantly stronger in this region and the transport is proportional to the cube of the velocity. The broad-scale transport is westward, with strong transport of 0.01-0.1 g cm-1 s-1 in the erosion or non-depositional environment near the eastern entrance and values of 10-4-10-3 g cm-1 s-1 in the transition zone from sand to mud toward the west. Bokuniewicz (1980) reported observed sand transport rates of 0.01 g cm-1 s-1, based on observations of sand-wave migration over 10 km scales in the east-central Sound. Superimposed on this westward transport is a strong clockwise transport of sand around the large, longitudinally directed sand ridges, particularly Long Sand Shoal. If we calculate the divergence of the bedload transport, we can determine the regions where we would expect sand to accumulate (convergent regions) and erode (divergent regions), provided there is ample sediment supply (Figure 14). There is a strong convergence of sand on most of these ridges, providing an explanation for their maintenance. On top of Long Sand Shoal, sand is predicted to converge at the rate of 0.011 cm day-1, or about 4 cm yr-1. At this rate, 20-m high ridges can form in 500 yr. This indicates that the tidal and density-driven flows provide a continuous mechanism for supplying sand to sustain the ridges, a process that is probably balanced by the dispersion of sand off the ridges due to surface waves during strong storm events. These simulated transport rates are regarded only as crude estimates, because bedload transport formulas are probably accurate within a factor of 2 or so (Van Rijn, 1984). Moreover, these simulations assume a constant grain size, and we have not considered variations caused by the spring/neap cycle or other sources of variability. Nonetheless, it is clear that the combination of tides and estuarine circulation is an effective transport mechanism that is consistent with the sea-floor sedimentary environment of the region.

fig. 13
Figure 13. Simulated bottom bedload transport driven by currents associated with the along-Sound density gradient and M2 tidal forcing. Arrows are not scaled to magnitude. Use color legend to determine magnitude of transport.

fig. 14
Figure 14. Divergence of simulated bedload transport shown in Figure 13. Regions of convergence (potential sand accumulation) are shown as shades of blue, and regions of divergence (potential sand erosion) are shown as shades of orange.

Conclusions

The results of this study provide a general description of the distribution and causes of bottom currents in Long Island Sound and their relationship to the sedimentary environments. In the funnel-shaped eastern part of the Sound, the gradient of tidal-current speed corresponds to a westward progression of sedimentary environments (Knebel et al., 1999; Knebel and Poppe, 2000). Currents here are sufficient to move sediments of fine sand and coarser and to produce coarse lag deposits in areas of erosion or nondeposition as well as winnowed finer sands in areas of bedload transport and sediment sorting. Although the tidal-current regime can explain most general aspects of the distribution of bottom environments, our modeling indicates that resuspension by waves is important along the nearshore margins of the Sound. In these shallow regions, the bottom-orbital speeds associated with surface waves are strong and are sufficient to resuspend fine-grained sediments (muds) about 1-10% of the time. The frequency of sediment movement drops dramatically with water depth, and waves have essentially no effect in water depths greater than about 20 m. Estuarine circulation and westerly wind events are shown to locally enhance tidal bottom currents along the axial depression of the Sound, providing a possible explanation for the relatively coarse sediments found at some locations in the depression. Simulated bedload-transport patterns, due to the superposition of tidal and estuarine flow, describe broad westward transport in the eastern Sound. The simulations also show strong transport around and convergence over the large sand ridges in the eastern Sound, consistent with their long-term maintenance. The complex patterns of simulated tide-induced residual circulation and bedload transport in the eastern Sound illustrate the difficulty in determining large-scale behavior from point measurements of currents and sediment profiles in regions of irregular bathymetry.

Acknowledgements

John Evans developed analysis and graphical tools that greatly facilitated this study. Ralph Lewis and Muriel Grim supplied us with bathymetry data that made construction of a high-resolution digital bathymetric grid possible. Thanks to Harley Knebel, Larry Poppe, and Jim O'Donnell for many helpful discussions.

Literature Cited

Blumberg, A.F. and Mellor, G.L., 1987. A description of a three-dimensional coastal model. In: N. Heaps (ed.), Three-Dimensional Coastal Ocean Models. Washington D.C.: American Geophysical Union, Coastal and Estuarine Sciences Series Number 4, pp.1-16.

Bokuniewicz, H.J., 1980. Sand transport at the floor of Long Island Sound. Advances in Geophysics, 22, 107-128.

Bokuniewicz, H.J. and Gordon, R.B., 1980a. Sediment transport and deposition in Long Island Sound. Advances in Geophysics, 22, 69-106.

Bokuniewicz, H.J. and Gordon, R.B., 1980b. Storm and tidal energy in Long Island Sound. Advances in Geophysics, 22, 41-67.

Bowman, M.J. and Esaias, W.E., 1981. Fronts, Stratification, and Mixing in Long Island and Block Island Sound. Journal of Geophysical Research, 86, 4260-4264.

Buchholtz ten Brink, M.R., Mecray, E.L. and Galvin, E.L., this volume. Clostridium perfringens in Long Island Sound sediments: An urban sedimentary record. Thematic Section. Journal of Coastal Research.

Csanady, G.T., 1973. Wind-induced barotropic motions in long lakes. Journal of Physical Oceanography, 3, 429-438.

Earwaker, K.L., 1990. Long Island Sound oceanography project: 1988-1990. NOS Oceanographic Circulation Survey Report No. 10, 40 p, with appendices.

Gadd, P.E., Lavelle, J.W., and Swift, D.J.P., 1978. Estimates of sand transport on the New York shelf using near-bottom current meter observations. Journal of Sedimentary Petrology, 48, 239-252.

Garvine, R. W., 1974. Physical features of the Connecticut River outflow during high discharge. Journal of Geophysical Research, 79, 831-846.

Gordon, R.B., 1980. The sedimentary system of Long Island Sound. Advances in Geophysics, 22, 1-40.

Gordon, R.B. and Pilbeam, C.C., 1975. Circulation in central Long Island Sound. Journal of Geophysical Research, 80, 414-422.

Grant, W.D. and Madsen, O.S., 1979. Combined wave and current interaction with a rough bottom. Journal of Geophysical Research, 84, 1797-1808.

Harris, P.T., Pattiaratchi, C.B., Cole, A.R., and Keene, J.B., 1992. Evolution of subtidal sandbanks in Moreton Bay, eastern Australia. Marine Geology, 103, 225-247.

Holthuijsen, L.H., Booij, N., and Herbers, T.H.C.; 1989. A prediction model for stationary, short-crested waves in shallow water with ambient currents. Coastal Engineering, 13, 23-54.

Hunter, J.R. and Hearn, C.J., 1987. Lateral and vertical variations in the wind-driven circulation in long, shallow lakes. Journal of Geophysical Research, 92, 13106-13114.

Ianneillo, J.P., 1981. Tidally-induced residual currents in Long Island and Block Island Sounds. Estuarine, Coastal and Shelf Science, 12, 177-191.

Jay, D.A. and Bowman, M.J., 1975. The physical oceanography and water quality of New York Harbor and western Long Island Sound. State University of New York at Stony Brook, New York. Marine Sciences Research Center, Technical Report #23. 71 p.

Kenefick, A.M., 1985. Barotropic M2 tides and tidal currents in Long Island Sound: A numerical model. Journal of Coastal Research, 1, 117-128.

Kim, B-H. and Bokuniewicz, H.J., 1991. Estimates of sediment fluxes in Long Island Sound. Estuaries, 14, 237-247.

Knebel, H.J. and Poppe, L.J., 2000. Sea-floor environments within Long Island Sound: A regional overview. Thematic Section. Journal of Coastal Research, 16(3), 535-550.

Knebel, H.J., Signell, R.P., Rendigs, R.R., Poppe, L.J., and List, J.H., 1999. Seafloor environments in the Long Island Sound estuarine system. Marine Geology, 155, 277-318.

Komar, P.D. and Miller, M.C., 1975. Sediment threshold under oscillatory waves. In: Proceedings, 14th Conference on Coastal Engineering. New York: American Society of Civil Engineers, pp. 756-775.

Lelacheur, E.A. and Sammons, J.C., 1932. Tides and currents in Long Island and Block Island Sounds. U.S. Coast and Geodetic Survey, Special Publication 174, 187 p.

Luettich, R.A., Jr. and Westerink, J.J., 1995. Continental shelf scale convergence studies with a barotropic tidal model, In: D. Lynch and A. Davies (eds.), Quantitative Skill Assessment for Coastal Ocean Models. Washington D.C.: American Geophysical Union, Coastal and Estuarine Studies Series 48, pp. 349-371.

Mecray, E.L. and Buchholtz ten Brink, M.R., 2000. Contaminant distribution and accumulation in the surface sediments of Long Island Sound. Thematic Section. Journal of Coastal Research, 16(3), 575-590.

Moody, J.A., Butman, B., Beardsley, R.C., Brown, W.S., Daifuku, P., Irish, J.D., Mayer, D.A., Mofjeld, H.O., Petrie, B., Ramp, S., Smith, P., and Wright, W.R., 1984. Atlas of tidal elevation and current observations on the northeast American continental shelf and slope. U.S. Geological Survey Bulletin 1611, 122 p.

Poppe, L.J., Knebel, H.J., Mlodzinska, Z.J., Hastings, M.E., and Seekins, B.A., 2000. Distribution of surficial sediment in Long Island Sound and adjacent waters: Texture and total organic carbon. Thematic Section. Journal of Coastal Research, 16(3), 567-574.

Redfield, A.C., 1980. The Tides of the Waters of New England and New York, Taunton, MA: William S. Sullwold Publishing, Inc., 108 p.

Rhoads, D.C., 1994. Analysis of the contribution of dredged material to sediment and contaminant fluxes in Long Island Sound. U.S. Army Corps of Engineers Special Technical Report 88, 32 p.

Riley, G.A., 1956. Oceanography of Long Island Sound, 1952-1954. Bulletin of the Bingham Oceanographic Collection, 15, 15-46.

Schmalz, R.A., 1993. Numerical decomposition of Eulerian residual circulation in Long Island Sound. In: M. Spaulding; K. Bedford; A.F. Blumberg; R. Cheng, and C. Swanson (eds.), Proceedings 3rd International Conference on Estuarine and Coastal Modeling. New York: American Society of Civil Engineers, pp. 294-308.

Schmalz, R.A., Devine, M.F., and Richardson, P.H., 1994. Residual Circulation and Thermohaline Structure. Long Island Sound Oceanography Project Summary Report, Volume 2, NOAA Technical Report NOS-OES-003. Rockville, MD: National Oceanic and Atmospheric Administration, 199 p.

Signell, R.P., Knebel, H.J., List, J.H., and Farris, A.S., 1998. Physical processes affecting the sedimentary environments of Long Island Sound. In: M. Spaulding and A.F. Blumberg (eds.), Proceedings 5th International Conference on Estuarine and Coastal Modeling, New York: American Society of Civil Engineers, pp. 400-412.

Smagorinsky, J., 1963. General circulation experiments with the primitive equations I. The basic experiment. Monthly Weather Review, 91, 99-164.

Speer, P.E. and Aubrey, D.G., 1985. A study of non-linear tidal propagation in shallow inlet/estuarine systems. Part 2: Theory. Estuarine, Coastal and Shelf Science, 21, 207-224.

Van Rijn, L.C., 1984. Sediment transport, part I: Bedload transport. Journal of Hydraulic Engineering, 110, 1431-1456.

Varekamp, J.C., Bucholtz ten Brink, M.R., and Mecray, E.L., 2000. Mercury in Long Island Sound sediments. Thematic Section. Journal of Coastal Research, 16(3), 613-626.

Wang, D-P., 1979. Wind-driven circulation in Chesapeake Bay, Winter 1975. Journal of Physical Oceanography, 9, 564-572.

Wilson, R.E., 1976. Gravitational circulation in Long Island Sound. Estuarine, Coastal and Shelf Science, 4, 443-453.

Wolfe, D.A., Monahan, R., Stacey, P.E., Farrow, D.R.G., and Robertson, A., 1991. Environmental quality of Long Island Sound: Assessment and management issues. Estuaries, 14, 224-236.

Back to the SMIG Features Page

button bar

SMIG Home Mailing List Features Conferences Classes Reading Room Model Archives Feedback Home | Mailing List | Features | Conferences | Classes | Reading | Model Archives | Feedback

Stewart Rounds, SMIG coordinator <sarounds@usgs.gov>
U.S. Geological Survey
http://smig.usgs.gov/SMIG/features_0900/li_sound.html
Last modified Wednesday, 17-Dec-2003 14:06:59 EST
Privacy Statement · Disclaimer · FOIA · Accessibility