3.1.2. Point Sources (CSOs, SWOs, and POTWs)

In addition to these freshwater tributaries, there are also significant anthropogenic freshwater inputs to Newark Bay, including urban runoff, CSOs, SWOs, and POTWs. Suszkowski [18] reported undifferentiated industrial and municipal wastewater discharges totaling 6.6 m3 /s (232 ft<sup>3</sup> /s) to Newark Bay. HydroQual Inc. [30] subsequently reported wastewater discharges during 1989, yielding a combined total of 13.4 m3 /s (474 ft<sup>3</sup> /s) of industrial and municipal runoff into tributaries of Newark Bay during that year. Another estimate places freshwater flows into the Arthur Kill from sewage treatment plants at 4.8 m3 /s or (170 ft<sup>3</sup> /s), and from permitted industrial discharges at 0.5 m3 /s (18 ft<sup>3</sup> /s) into the Arthur Kill [29]. A more recent approximation indicates that the Bergen County POTW and Secaucus POTW discharge an average of 3 m3 /s (108 ft<sup>3</sup> /s) and 0.15 m3 /s (5.3 ft3 /s), respectively, to the Hackensack River [31].

**Figure 2.** Freshwater flows into the NBSA where the size of the open arrows reflects the relative magnitude of the flows.

#### 3.1.3. Non-Point Sources (Direct Runoff and Groundwater)

The State of New Jersey's annual rainfall typically ranges between 813 mm (32 in.) and 1219 mm (48 in.) with an average of 1151 mm (45.3 in.) from 1895 to 2012 [32]. Newark averages slightly more rain at 1174.8 mm/year (46.25 in./year) [33]. Most of the rain falling on the Newark Bay watershed eventually enters the Bay in the form of runoff or groundwater influx, the amount of which is highly variable and dependent on the annual climatic conditions.

### 3.1.4. Flows through the Tidal Straits

Newark Bay is connected to the Atlantic Ocean by two straits: the Kill van Kull on the southeast and the Arthur Kill on the southwest. The Kill van Kull connects Newark Bay to the Upper Bay of New York Harbor (to the east), while the Arthur Kill connects Newark Bay to Raritan Bay (to the south). The combination of tidal forcing, significant freshwater flows from the Passaic and Hackensack Rivers, and dredged navigational channels results in measurable salinity stratification within both tidal straits, yielding daily-averaged, two-layer subtidal flow patterns with seaward flows of less saline water near the surface that are offset by landward flows of saltier water near the bottom [16,22,34].

The Kill van Kull is a tidally dominated strait where the tidal excursion (distance that water travels over half a tidal cycle) is much greater than its length. Because of its relatively short length, the Kill van Kull exhibits strong currents and relatively weak stratification (up to 1.5 psu), promoting a greater degree of tidal exchange and mixing between Newark Bay and New York Harbor than occurs between Newark Bay and Raritan Bay by way of the Arthur Kill [16,35]. The movement of water through the Arthur Kill is impeded relative to the Kill van Kull, because the tidal excursion is shorter than its length. Its freshwater sources support a mild vertical salinity gradient of up to 1 psu. LMS [36] estimated tidal flows through the Kill van Kull and the Arthur Kill to occasionally exceed 1417 m3 /s (50,000 ft3 /s) and 283 m3 /s (10,000 ft<sup>3</sup> /s), respectively, with average flows passing through Kill van Kull ranging up to an order of magnitude larger than those passing through Arthur Kill [23].

#### 3.1.5. Gravitational Circulation and Tidal Currents

Flows through the tidal straits are a result of both salinity gradients (gravitational circulation) and water-elevation differences (tidal current) between the ends of each strait (meteorological forcing also plays an important role). Freshwater inflow, primarily from the Passaic and Hackensack Rivers, supports both the gravitational circulation and tidal current mechanisms. Figure 3 is a schematic of the two-layer circulation in the NBSA. Red arrows indicate salt water flow, and blue arrows freshwater flow. For the relatively larger inputs, the numbers near each arrow indicate the volumetric inflow over a tidal cycle as a fraction of the estimated volume of the Bay (with water at sea level). The Kill van Kull and Arthur Kill straits tend to form a through-flow pattern of circulation around Staten Island, in contrast to a tidal pumping mode of circulation in which Newark Bay is filled (or emptied) by simultaneous inflow (or outflow) through both Kills [16,19,20].

**Figure 3.** Schematic of the two-layer circulation pattern in the NBSA.

#### 3.1.6. Meteorological Forcing

Although tidal circulation patterns through the Kill van Kull and the Arthur Kill are generally counterclockwise around Staten Island [20], strong and persistent meteorological events can, at times, dominate the circulation pattern in Newark Bay, as well as the magnitude and direction of flows in the Kill van Kull and the Arthur Kill [16,17,29,37,38]. The primary types of storms that affect Newark Bay include tropical storms that typically occur in late summer and fall, and extra-tropical ("Nor'easter") storms that occur primarily in the winter. The extra-tropical storms can cause high water levels and enhanced wave conditions. Potential wave heights can be over 1.8 m (6 ft) for the most severe storms, but are typically less than 1.2 m (4 ft) [39]. Wind-generated currents affect water mixing and possibly sediment transport within Newark Bay [39]. Strong wind events were also shown to generate large episodic flushing.

#### *3.2. Sediment Dynamics*

Sediment dynamics in estuarine environments such as the NBSA are driven by several factors, including hydrodynamics, episodic meteorological events, sediment loading from freshwater inflows, sediment loading at open boundaries, sediment size gradation, bed sediment properties, bioturbation, and particle-to-particle interactions. Significant human activities including long term maintenance and navigational dredging and ship traffic also affect sediment dynamics. A significant volume of research has been conducted to better understand the role that tidal circulation patterns and wind-driven episodic events play in the fate and transport of sediments within the NBSA [15,16,18,19,22,34,37,38,40]. Fluid mud transport is not considered in this CSM because site characterization data collected to date have not indicated the presence of fluid mud. In addition, no previous publication related to the NBSA has considered the effects of fluid mud as a potential transport pathway for sediments and associated contaminants. Until such processes are observed, there is no basis to include them in the CSM.

#### 3.2.1. Sediment Transport Processes

A schematic of the processes that influence sediment transport in the water column and the sediment bed of the NBSA is shown on Figure 4. Local advection and dispersion in the water column control the distribution of sediment particles throughout the system. Advection moves the sediments according to the local water velocity while dispersion spreads sediments based on concentration gradients. Sediments are typically classified as either cohesive (small grain sizes) or noncohesive (larger grain sizes). Cohesive sediments are composed primarily of clay-sized (<2 ȝm) and silt-sized (<63 ȝm) particles, mixed with organic matter and sometimes small quantities of very fine sand. Noncohesive sediments are primarily sand and gravel-sized materials (>63 ȝm). Each sediment class is subject to different physical processes [41–43].

### 3.2.2. Bed Shear Stresses

Hydrodynamic flows will result in variable shear stresses at the sediment bed that, depending upon erodibility, may lead to erosion. When wind-waves are present, it is necessary to account for the shear stress on the sediment bed caused by wave-current interaction, which is a function of the bottom orbital amplitude and bottom orbital currents, both of which depend on the wave climate [44,45]. Including the effect of waves is necessary because the bed shear stresses can be an order of magnitude greater than stresses caused by currents alone. Vessel-generated wakes, associated with tugs, barges, and deep-draft vessel traffic in the Navigation Channels, represent another source of wave action in the Bay [10,15]. Although minor compared to wind-waves, these waves can also contribute to the resuspension of bottom sediments in shallow Subtidal Flat areas. Ship traffic can also resuspend sediments in the deeper dredged portions of the channel due to prop wash.

**Figure 4.** Schematic of sediment transport processes where *C* is the suspended sediment concentration profile, *U* is the velocity profile, *IJ<sup>b</sup>* is the bed shear stress, and *u\** is the shear velocity. Concentration and velocity profiles are conceptual only. The concentration profile reflects increased sediment concentration with depth, and the velocity profile decreases to zero at the sediment bed.

3.2.3. Sediment Loading

In terms of general sediment inputs to the NBSA, LMS [36] estimated that the Kill van Kull and the Arthur Kill provide approximately 60% and 12% of the total sediment load to the Bay, respectively. The Passaic and Hackensack Rivers deliver 23% and 2% of the total sediment load, respectively. Other sources of sediment to Newark Bay include POTWs (Bergen County and Secaucus), CSOs, SWOs, and atmospheric deposition, which are estimated to deliver a combined total of approximately 3% of the total sediment load [36]. Figure 5 shows the relative magnitude of the sediment loading to the NBSA (scaled by the percentages listed above). These estimates are similar to those of Suszkowski [18], Lowe *et al.* [31], and Sommerfield and Chant [22]. The two notable differences are that Sommerfield and Chant [22] and Pence [17] suggest that the Arthur Kill is a net exporter of sediment, and that Suszkowski [18] found the Hackensack to be a net sink for Bay sediments. There is significant variability across the estimates, the consequence of which is that there is appreciable uncertainty in the annual sediment input to Newark Bay. Table 1 lists published sediment loading estimates for the Passaic River into Newark Bay.

**Figure 5.** Sediment loads to the NBSA where the size of the arrows reflects the relative magnitude of sediment loading.

Sediment transport modeling conducted by Wakeman III [15] and sediment transport observations by Sommerfield and Chant [22] suggest that suspended sediments in the upper portion of the Bay do not leave the Bay during ebb tide or during periods of normal freshwater discharge to the Bay, and only up to about 15% of sediments in the lower portion of the Bay might exit the system. Sommerfield and Chant [22] evaluated suspended sediment deposition patterns during a Passaic River high-flow event and found that sediment deposition from such an event was greatest in the northern portions of the Bay, primarily within the Navigation Channel, with little evidence of flood-tide-related deposition on the Subtidal Flats. In the southern portion of the Bay, Sommerfield and Chant [22] indicate that most of the Kill van Kull sediment influx is not carried into the northern portion of the Bay. This suggests that Newark Bay experiences a localized convergence of sediment flux (deposition), which is consistent with the hypothesis [22] that the historical dredging of the Bay is required because otherwise the system tends to return to its natural, shallow state.


**Table 1.** Passaic River sediment loading rates.

1 Based on measured sediment flux data; 2 Based on a solids balance of sediment loads through the Dundee Dam from Lowe *et al.* [31] with deposition of 2.54 cm/year (1 in./year) along the Passaic River; 3 Based on the chemical mass balances of 2,3,7,8-tetrachlorinated-*p*-dibenzodioxin (TCDD) and total TCDD of the solids mass balance in note 1; 4 Based on a sediment yield of 39 MT/km2 /year for the Passaic River watershed; 5 Based on measured sediment flux data; however, it is not clear if these are tons/year or MT/year.

#### 3.2.4. Historical and Ongoing Dredging Activities

As documented in USACE [16], the volume of the Navigation Channels excavated in Newark Bay has continually grown since the early 1900s. Moreover, substantial dredging is required simply for maintenance of the existing channels.

Olsen *et al.* [48] reported that the average annual dredge volume (Newark Bay, the Kills, and the Passaic and Hackensack Rivers) measured by the USACE in 1942–1973 was 439,864 m3 /year (575,320 yard3 /year; 220,000 MT/year assuming 0.5 g/cm3 dry density). This volumetric rate is higher than that reported by Lowe *et al.* [31], who evaluated USACE data from 1924 to 1985 and reported that the average annual dredge volume for 1953–1985 (when the channels were fairly stable) was 161,680 m3 /year (211,469 yard3 /year or 80,840 MT/year). Moreover, Lowe *et al.* [31] also reported estimates from the Port Authority of New York and New Jersey, because they dredged a significant additional amount of 100,388 m3 /year (131,303 yard3 /year or 50,194 MT/year). After studying several sources of information, Wakeman III [15] concluded that the majority of the annual sediment load of 276,000 m3 /year (361,000 yard3 /year or 138,000 MT/year) is being removed by maintenance dredging operations in the USACE channels and private berths totaling 262,000 m3 /year (342,683 yard3 /year or 131,000 MT/year). It is the general consensus that the Navigation Channels in Newark Bay are the ultimate sinks for most fine-grained sediments entering the Bay, while the tidal flats are only temporary repositories for sediments that are subsequently resuspended for deposition into the Navigation Channels or for export.

More recently, the Port Authority of New York and New Jersey's Harbor Deepening Project (HDP) included dredging the channels from the Ambrose Channel entrance to the Upper Bay and Newark Bay, providing access to the Global Marine Terminal, New York Container Terminal, Port Newark, and Elizabeth Marine Terminal. Over 2,752,397 m3 (3,600,000 yard3 ) will be dredged by 2014. In 2011, 405,979 m3 (531,000 yard3 ) was dredged from Newark Bay. By 2013, 298,940 m3 (391,000 yard3 ) of silt and 1,070,377 m3 (1,400,000 yard3 ) of clay, sand, and blasted rock were dredged from the Arthur Kill. The HDP channels constitute 21% of the total area in Newark Bay. In channels that are deep and flat, the sedimentation rate is moderate at 3 cm/year (0.1 ft/year). USACE [10] modeling suggests that the HDP will only have small effects on sedimentation on the flats because the planned dredging will not change the configuration of the channels—it will only deepen the existing channels. Sommerfield and Chant [22] and a modeling study by Pecchioli *et al.* [19] suggest increased sediment deposition in the Bay due to channel deepening at the Kill van Kull and Arthur Kill.

#### 3.2.5. Overall Sediment Dynamics in Newark Bay

Particle size, salinity, and velocity gradients are key factors in sediment transport within the NBSA. Burke *et al.* [40] indicate that the Navigation Channels act as the primary pathway for sediment transport and, once suspended, the fate of the sediments in the Navigation Channel depends on many factors, including the size of the particles and their settling velocity. Heavier particles tend to settle more quickly into the Navigation Channel bed, while finer particles that remain suspended during flood tide are caught in the gravitational estuarine circulation and transported to the northern portion of the Bay. During ebb tide, these same particles tend to settle and, depending upon conditions, deposit onto the sediment bed [15].

Water-column stratification also has important implications for sediment transport. During ebb tides, stratification is intensified, significantly reducing resuspension and encouraging suspended sediment deposition, particularly in the Navigation Channel [22]. For example, during a 2001 high-flow event on the Passaic River, Chant [16] reported increased suspended sediment concentrations within Newark Bay, and that coarser particles settled out of the seaward-flowing surface water into the landward bottom flow, effectively becoming trapped. Sommerfield and Chant [22] found that sediment deposition from a high-flow event in the Passaic River was greatest in the northern portions of the Bay, with little evidence of flood-tide-related deposition on the Subtidal Flats, and that most of the sediment influx from the Kill van Kull was not carried into the northern portion of the Bay, which led them to conclude that greater Newark Bay acts as a

sediment convergence zone. This process corresponds to the dredged Bay bathymetry moving toward equilibrium of a natural, shallow state. Moreover, sediment may be deposited preferentially near and along the base of the steeply sloped edges of the Navigation Channel. Concurrently, the steep banks of the Navigation Channels may be eroded preferentially at the uphill edge with sediment transported downslope and into the toe of the Navigation Channel [10].

Sommerfield and Chant [22] also observed a short-term convergent deposition pattern based on their analysis of Be-7 in the surface sediments. There was a large range in Be-7 inventory, 0.2 to 6.7 pCi/cm2 , with higher Be-7 inventories detected in the Navigation Channels than in Subtidal Flats. The Be-7 stations in the Navigation Channel also appeared to be responsive to a Passaic River high-flow event, with a sharp increase in inventory. In the Navigation Channels of the northwestern portion of the Bay (near the mouth of the Passaic River) and around Shooters Island in the south, Be-7 was detected to a depth of 2.4 in. In the Subtidal Flats, however, a much thinner sediment layer, less than 0.8 inches, was found to have Be-7 activity. The differences in Be-7 depth are thought to represent differences in the physical mixing present in the various areas. Seasonal deposition and bed reworking appear to be relatively intense in the Navigation Channels of the northwestern portion of the Bay, as well as the southern portion of the Bay around Shooters Island (where tidal currents and vessel-induced stresses are strong) compared to the shallower Subtidal Flats where biological mixing dominates [22].
