Reconstructing Contaminant Deposition in a San Francisco Bay Marina, California

Abstract

Two sediment cores were collected from a marina in the San Francisco Bay to characterize historical sediment contamination resulting from the direct discharge of industrial wastewater from Naval Air Station Alameda. Depth profiles of trace metals, petroleum hydrocarbons, and radionuclides were determined with a 12-cm spacing down to a depth of 120 cm. The chronology of sediment accumulation is established by depth profiles of sedimentary time markers in conjunction with information on site history. The traditional approach of determining sediment accumulation rates by measuring atmospheric ²¹⁰Pb deposition was obscured by a larger source of ²¹⁰Pb in the sediments from the decay of anthropogenic ²²⁶Ra, likely from luminescent paints used at this facility and released to the marina. The sedimentation rates inferred from the data indicate that the greatest amount of contamination by trace metals and petroleum hydrocarbons took place between 1940 and 1960. In addition, anthropogenic ²²⁶Ra activities are positively correlated with some of the contaminants in the sediments, allowing the wastewater discharged from the facility to be distinguished from baywide contamination. In locations such as this, where there is a complex history of contaminant deposition, a source-specific tracer may be the only feasible method of attributing historical contamination to a point source.

Introduction

Contaminated marine sediments are an important international problem that requires appropriate methods for site characterization and remediation. Contaminated sediments are typically found in estuaries near urban, industrial, and agricultural areas where fine-grained sediments accumulate. In general, the spatial extent of contamination is extensive while the vertical extent is limited. Two issues drive the need for characterization and remediation of contaminated sediment: the adverse impact on the ecosystem and the potential for contaminant bioaccumulation into fish and shell-fish used for human consumption National Research Council (NRC 1997).

The San Francisco Bay has received wastes from municipal, industrial, agricultural, and mining activities that have altered the sediment quality (Hornberger et al. 1999). A recent study of the impact of human activities on sediments within San Francisco Bay (van Geen and Luoma 1999) demonstrated that historical contamination in Northern California was recorded chronologically in the sedimentary record. If sediment remediation is warranted, the contaminant depositional record must be deconvoluted into contaminants from both regional nonpoint sources and local point sources. This study characterizes the vertical distribution of contaminants in two sediment cores from a marina in the San Francisco Bay that is adjacent to industrial wastewater outfalls. Sedimentation rates in the marina are determined using natural and anthropogenic radionuclides ²¹⁰Pb and ¹³⁷Cs along with historical records. These rates are then used to reconstruct contaminant depositional histories. The contribution of contamination from local sources is distinguished from baywide contamination by identifying and correlating a component specific to local industrial operations. This study was part of a larger environmental investigation by the U.S. Navy to close the air station and return the facility to the local community (PRC Environmental Management, Inc. 1994; O’Day et al. 2000).

Krishnaswami et al. (1971), Koide et al. (1973), and Robbins and Edgington (1975) originally established the use of ²¹⁰Pb and ¹³⁷Cs to measure sedimentation rates in marine and lacustrine sediments. This technique was promptly applied to industrial pollution in the sedimentary record (Bruland et al. 1974; Edgington and Robbins 1976; Schubel and Hirschberg 1977; Goldberg et al. 1979). Many recent studies have incorporated this geochronological method to constrain the time of contaminant deposition (Bopp and Simpson 1989; Macdonald et al. 1991; Valette-Silver 1993; Krom et al. 1994; Gerritse et al. 1998; Palanques et al. 1998; Smith and Schafer 1999; van Geen and Luoma 1999; Yunker et al. 1999), but none distinguished specific point sources in complex depositional environments.

Site History

The former Alameda Naval Air Station is one of many military bases that has closed and requires a site investigation prior to returning the facility to the local community. The Seaplane Lagoon is a marina within this facility at the western end of Alameda Island, adjacent to the city of Oakland as shown in Figs. 1(a and b). The rectangular marina was constructed in the 1930s on mudflats by the use of seawalls and dredging to a water depth of 6–7 m that reached into a consolidated sand formation called the Merritt Sand. The surface area of the lagoon is 4.5×10⁵ m² and there is a 250-m-wide opening with access to San Francisco Bay. The marina is a site for sediment accumulation since it was dredged deeper than its preconstruction condition and is relatively protected from the currents and wind-generated waves in the larger bay system. Even so, there was no maintenance dredging required during facility operations. During the period 1943–1975 storm water and untreated industrial wastewater exclusively from the air station were discharged into the marina through seven outfalls indicated in Fig. 1(b). Use of ²²⁶Ra to enhance night visibility of indicator needles, knobs, and deck markings began in the 1940s in the dial paint section of the instrument shop located in Building 5 treatment [Naval Energy and Environmental Support Activity (NEESA 1983)]. Normal operations in this building involved scraping old paint, cleaning with solvent, and repainting with paint containing ²²⁶Ra. Wastes from this location were discharged directly into the Seaplane Lagoon through the western outfall locations. The radium dial paint shop was closed in the late 1950s or early 1960s. Any refurbishing of radioluminescent paint was then performed on the second floor of Building 400 treatment (NEESA 1983). Between 1972 and 1975 the industrial wastewater from the whole facility was isolated and diverted for treatment (NEESA 1983). The industrial wastewater plumbing system from Buildings 5 and 400 has been removed due to high ²²⁶Ra levels on the interior pipe surfaces.

Fig. 1.

(a) Regional location of Naval Air Station Alameda and (b) diagram of Seaplane Lagoon. Arrows indicate location of wastewater outfalls. Star indicates sampling location. Relevant buildings are indicated by their building numbers

Experimental Methods

Coring

In July 1997, two sediment cores were collected using a gravity corer near the wastewater outfalls at the northeast corner of the marina as indicated in Fig. 1. These two cores were separated by approximately 3 m. The gravity corer was 1.8 m long with a clear acrylic liner (9 cm internal diameter) and weighed 330 kg. Each sediment core was approximately 120 cm in length. Complete cores were collected because the sediment-water interface on the inside was the same as the mud level on the outside of the core barrel. In addition, the water above the sediment was not turbid, indicating no sediment loss by resuspension. Cores were maintained in a vertical position at all times during core acquisition and transport to shore. The cores were packed vertically in dry ice for 24 h until completely frozen to retain any semivolatile organic constituents and to prevent disturbance of the core during transport to the laboratory. Once in the laboratory, the cores were stored at −15°C until they were sampled.

Both frozen cores were sectioned with a hacksaw into ten intervals, each approximately 12 cm deep. One core was used for radiologic analysis and bulk density determinations; the other was subsectioned for organic, metal, and bulk density determinations. Samples subsectioned for organic and metal analyses were approximately 9 cm in length; samples subsectioned for radiologic analyses were approximately 5 cm long; and bulk density profiles used subsections of approximately 2.5 cm. The frozen sediment was then extruded from the acrylic liner by hand and trimmed with a stainless steel surgical saw on all sides to avoid contamination from smearing during coring, sawing, and extrusion.

Dry bulk density was determined by gravimetric analysis after heating at 104°C. Porosity was determined from the measured wet volume and mass loss of the wet sediment after drying, assuming all pore space was filled with water.

Chemical Analysis

Samples for organic and metals analyses were kept frozen until they were thawed for extraction and analyses. The samples were homogenized over the depth interval of each sample. Methods for determining organic and metal concentrations in the sediment followed EPA guidelines (EPA 1996). The organic contaminants were extracted ultrasonically from the thawed sediment (3550A method). Total petroleum hydrocarbons (TPH) were measured using gas chromatography (8100 modified method). TPH were calculated from the gas chromatogram using Diesel No. 2 as the standard that most closely matched the hydrocarbon fingerprints in the samples. The metal contaminants were extracted from the thawed sediment by acid digestion (3050 method). The metals were measured using inductively coupled plasma–mass spectrometry (6020 method).

Radionuclide Analysis

The activities of the radionuclides ⁴⁰K, ¹³⁷Cs, ²²⁶Ra, ²²⁸Ra, and ²²⁸Th were determined by gamma spectrometry using germanium solid-state detectors with a planar geometry. Details of the method are described by Wong et al. (1992); Bandong et al. (2001); and Volpe et al. (2001). ⁴⁰K was quantified due to its association with clay minerals, ¹³⁷Cs as a tracer of atmospheric weapons testing, and radium and thorium isotopes to examine natural and anthropogenic sources of other radionuclides in the sediment. A 60–100 g sample of freeze-dried sediment was hermetically sealed into an aluminum can and stored for 30–60 days to allow ²²²Rn (3.8-day half-life) and ²¹⁴Pb (26.8-min half-life) to establish secular equilibrium with ²²⁶Ra. Secular equilibrium occurs when the activities in a radionuclide decay series are equal. Samples were counted for 2–3 days in a low-level counting facility at Lawrence Livermore National Laboratory (LLNL) on calibrated detectors. Uncertainties (1 SD) from counting statistics, peak fitting, and line averaging were between 1 and 5%. With the exception of ¹³⁷Cs, the nuclides mentioned were seen in every sample. Detection limits for ¹³⁷Cs ranged from 8.3×10⁻⁵ to 4.3 ×10⁻⁴ Bq/g (0.005–0.026 dpm/g, where dpm indicates disintegrations per minute).

²¹⁰Pb activity was determined by isotope-dilution alpha spectrometry of its ²¹⁰Po granddaughter. The details of the method are described by Benoit and Hemond (1988); Wong et al. (1992); and Ritson et al. (1994). The method assumes secular equilibrium between ²¹⁰Pb and ²¹⁰Po, which is reasonable for sediments. In preparation of marine sediments for analysis, a standard practice is to leach approximately 1 g of dry sediment in hot nitric acid after addition of a ²⁰⁹Po tracer, evaporate the leachate to dryness, reconstitute the residue in hydrochloric acid, and then deposit polonium onto a small silver planchet. The planchet is then alpha counted using partially depleted silicon surface barrier detectors, and the ²¹⁰Po activity is determined from the measured ²¹⁰Po/²⁰⁹Po activity ratio. The uncertainty for the measurements is less than 6% (1 SD) and is dominated by counting statistics.

The normal technique for ²¹⁰Pb determination was modified to remove the high concentration of refractory organic matter. One set of samples was dry ashed in a muffle furnace at 450°C (Krom and Berner 1983) and then leached in hot 1:3 hydrochloric:nitric acid. A second set of samples was leached in 1:3 hydrochloric:nitric acid and then wet ashed with hydrogen peroxide in concentrated nitric acid. Studies at LLNL (Wong et al. 1992) have shown that hot acid leaches quantitatively extract ²¹⁰Po from marine sediments. Results from samples dry ashed were systematically low, indicating some loss of ²¹⁰Po; therefore only ²¹⁰Pb data from the wet-ashing protocol were considered.

Results

Seaplane Lagoon sediment properties are not uniform throughout the recovered depth of 120 cm. The sediment was black fine-grained silty clay from 0 to approximately 90 cm below the sediment-water interface. From 90 cm to the bottom of the core the sediment rapidly graded into a gray silty sand of the Merritt Sand formation. There was no visual evidence of sediment reworking by benthic grazing organisms or benthic infauna. Bulk density and porosity are reported in Tables 1(a and b) for the two cores. Over the depth interval 0–90 cm below the sediment-water interface the dry bulk density is about 0.4 g/cm³ and the porosity is about 75% in both cores. Since there is no evidence for compaction of the sediment over the depth interval of 0–90 cm, sedimentation rates are calculated as depth per time instead of reporting mass accumulation rates. The physical properties of the sediment were disturbed below 90 cm by the core retainer used to keep the sediment in the acrylic liner during sampling. Changes in sediment mineralogy can be discerned through ⁴⁰K measurements since ⁴⁰K is often used as a surrogate for clay content. The ⁴⁰K activity at each measured interval (Table 2) from 0 to 90 cm below the sediment-water interface was 0.45±0.05 Bq/g (27±3 dpm/g). The uniformity in ⁴⁰K activity and physical properties of the sediment indicates that there was no substantial change in sediment composition during the time period that deposited the upper 90 cm of sediment.

Table 1.

Sediment Bulk Properties

Average depth (cm)	Sample interval (cm)	Dry bulk density (g/cm3)	Porosity (–)
(a) Isotope core
6.4	3.1	0.41	0.75
19	2.9	0.40	0.74
32	2.7	0.37	0.74
45	2.6	0.34	0.76
57	2.6	0.34	0.74
70	2.8	0.34	0.76
83	2.3	0.48	0.74
95	2.8	0.93	0.57
108	2.8	0.53	0.70
121	2.6	0.68	0.66
(b) Organic and metals core
10	2.6	0.42	0.74
23	2.4	0.43	0.73
33	2.9	0.45	0.74
45	2.8	0.36	0.76
56	2.3	0.40	0.76
67	3.1	0.36	0.73
79	2.2	0.36	0.78
90	2.4	0.49	0.74
102	2.7	0.95	0.48
113	2.3	0.80	0.76

Table 2.

Sediment Radionuclide Levels

Average depth (cm)	Sample interval (cm)	40K (dpm/g)	137Csa (dpm/g)	226Raa (dpm/g)	228Raa (dpm/g)	228Tha (dpm/g)
2.5	5	30.6	0.58	0.91	1.14	1.10
15	5	26.1	0.65	0.87	1.18	1.25
28	5	31.0	1.10	1.14	1.30	1.32
41	5	28.0	0.83	1.41	1.47	1.45
53	5	22.7	0.09	2.73	1.45	1.48
66	5	24.8	LT	3.32	1.41	1.48
79	5	25.8	LT	1.65	1.39	1.44
91	5	23.9	LT	1.33	1.18	1.29
104	5	31.1	LT	1.11	1.28	1.32
117	5	24.0	LT	1.12	1.29	1.35

Note: LT indicates below detection limit.

^aNormalized to ⁴⁰K variation

Chemical contaminants are not uniformly distributed over the sediment profile. Concentrations of five metals were measured within ten intervals and are plotted in Fig. 2 and included in Table 3. There were significant increases above surface sediment levels in Cd, Cr, and Pb from 39 to 85 cm below the sediment-water interface. Within this interval, Cd increased 14-fold, Cr increased two- to threefold, and Pb increased threefold above surface sediment levels. The concentration of Cu and Zn increased slightly above surface sediment levels at depths of 35–65 cm. The concentration of Ni was below the quantitation limit of 85 mg/kg at every depth, while all other metals were significantly above the quantitation limit. The concentration of total petroleum hydrocarbons (Fig. 3 and Table 3) in the sediment varies from about 800 mg/kg in near surface sediments to levels as high as 4,200 mg/kg between the depths of 50 and 85 cm below the sediment-water interface.

Fig. 2.

Depth profile of metal concentrations in Seaplane Lagoon sediment. Multiple points for each metal at 27 and 96 cm below sediment-water interface indicate duplicate measurements.

Table 3.

Sediment Metals and Total Petroleum Hydrocarbon Concentrations

Average depth (cm)	Sample interval (cm)	Cd (mg/kg)	Cr (mg/kg)	Cu (mg/kg)	Pb (mg/kg)	Zn (mg/kg)	Ni (mg/kg)	Total petroleum hydrocarbons (mg/kg)
4.5	9	10.9	124	89	230	230	LT	1,130
16	9	15.0	139	98	230	240	LT	830
27	9	16.2	125	80	280	197	LT	820
		17.5	127	81	320	210	LT
39	9	43	270	143	630	340	LT	900
50	9	77	220	106	600	260	LT	2,800
62	9	153	400	125	620	380	LT	3,600
								4,200
73	9	119	230	76	500	178	LT	3,000
85	9	115	210	81	260	195	LT	2,000
96	9	11.3	113	45	147	138	LT	1,200
		9.9	114	50	152	163	LT	1,300
107	9	13.6	111	56	136	156	LT	1,900

Note: LT indicates below detection limit.

Fig. 3.

Depth profile of total petroleum hydrocarbons in Seaplane Lagoon sediment. Multiple points for TPH at 62 and 96 cm indicate duplicate measurements.

The activity profile of gamma-emitting radionuclides is shown in Fig. 4 and the data are presented in Table 2. The activities of ¹³⁷Cs, ²²⁶Ra, ²²⁸Ra, and ²²⁸Th are normalized to the ⁴⁰K variation using the following equation:

$$A_{\text{norm}}=A_{\left(z\right)}\times A_{{}^{40}\mathrm{K}\left(\mathrm{avg}\right)}∕A_{{}^{40}\mathrm{K}\left(z\right)}$$

This normalization accounts for variations in radionuclide abundance associated with changes in clay content.

Fig. 4.

Depth profile of gamma-emitting radioisotopes in Seaplane Lagoon sediment. Measurements normalized to ⁴⁰K variation.

¹³⁷Cs is an anthropogenic radioactive isotope introduced into the global environment from fission reactions by atmospheric nuclear weapon testing. ¹³⁷Cs has been deposited from the atmosphere into the sedimentary record since the inception of such activity in 1952 and fallout of ¹³⁷Cs peaked in 1963. Since the position of the ¹³⁷Cs maximum is at a sediment depth of 28±6 cm, this depth is a chronological marker for the year 1963 and suggests an average sedimentation rate of 0.8±0.2 cm/year from 1963 to 1997. The 6-cm uncertainty in the peak depth of ¹³⁷Cs activity is attained from half the distance between sediment samples. The surface and near-surface ¹³⁷Cs activities are comparable to activities measured elsewhere in the Bay (Fuller 1982; Fuller et al. 1999).

The radionuclides ²²⁶Ra and ²¹⁰Pb are both constituents of the ²³⁸U decay series

$$\begin{array}{cc}\hfill {}^{238}\mathrm{U}& \to {}^{234}\mathrm{U}\to {}^{230}\mathrm{Th}\to {}^{226}\mathbf{Ra}\to {}^{222}\mathrm{Rn}\to {}^{218}\mathrm{Po}\to {}^{214}\mathrm{Pb}\hfill \\ \hfill & \to {}^{214}\mathrm{Po}\to {}^{210}\mathbf{Pb}\to {}^{210}\mathrm{Bi}\to {}^{210}\mathrm{Po}\to {}^{206}\mathrm{Pb}\hfill \end{array}$$

The pronounced ²²⁶Ra maximum at 66 cm depth is anomalous compared to other San Francisco Bay sediment profiles (Fuller et al. 1999). Over the sediment depth intervals of 0–28 cm and 91–120 cm, the ²²⁶Ra activities are relatively constant and are similar to ²²⁶Ra activities found in other Bay sediments (Fuller et al. 1999). There are three possible reasons for enriched ²²⁶Ra within the core: (1) this sediment layer is mineralogically enriched in radium, (2) the sediment layer is mineralogically enriched in ²³⁸U, which then decays to ²²⁶Ra, or (3) there is an anthropogenic source of ²²⁶Ra.

The reason for ²²⁶Ra enrichment within the core is explained by examining other radionuclides over the sediment profile. If the sediment minerals are enriched in radium overall, the isotope ²²⁸Ra should also be enriched, but the ²²⁸Ra profile is uniform over the depth of the core. Thus, there is no enrichment in radium. The uniform composition of ²²⁸Ra and ²²⁸Th in the sediment also argues against an enriched U sediment source, since sediments enriched in U are often also enriched in Th because of geochemical similarities (Faure 1986). If U were the source for ²²⁶Ra, than these radioisotopes would be enriched in the sediment profile where ²²⁶Ra is elevated. Since neither a natural radium nor uranium source is indicated, the most likely source of the excess ²²⁶Ra in the Seaplane Lagoon sediment is anthropogenic. The historical record of ²²⁶Ra-based painting at this facility supports this analysis.

Sediment accumulation rates during the last 10–100 years are commonly determined by the ²¹⁰Pb method. This method uses the natural deposition of ²¹⁰Pb, but is complicated by multiple ²¹⁰Pb sources in the sediment. In a closed system, all of the ²¹⁰Pb results from the ²³⁸U decay series. In contrast, sediments accumulate ²¹⁰Pb from atmospheric and water column sources as well as production within the sediment. The ²¹⁰Pb produced within the sediment is called “supported” and the ²¹⁰Pb from other sources is called unsupported or “excess.” The atmospheric source of ²¹⁰Pb is from the decay of ²²²Rn and the water column contributes ²¹⁰Pb from the decay of dissolved ²²⁶Ra. These additional ²¹⁰Pb contributions to the sediments from the atmosphere and water column provide the excess ²¹⁰Pb. Supported ²¹⁰Pb is generally determined by measuring ²²⁶Ra activity in the core, assuming secular equilibrium, or by determining ²¹⁰Pb activity at depth, where all excess ²¹⁰Pb has decayed away. The excess activity of ²¹⁰Pb is determined by subtracting the supported ²¹⁰Pb from the total measured ²¹⁰Pb. If the flux of atmospheric and water column ²¹⁰Pb and the sedimentation rate are constant over time, a log-linear plot of excess ²¹⁰Pb against depth in the sediment will yield a straight line as the excess ²¹⁰Pb decays exponentially. The anthropogenic radium in the Seaplane Lagoon, a nonmineralogic contributor of ²¹⁰Pb that presumably comes from World War II luminescent paint operations, obscures the excess ²¹⁰Pb values by contributing ²¹⁰Pb produced by the deposited paint. As a result, the atmospheric and water column excess ²¹⁰Pb cannot be adequately distinguished from the ²¹⁰Pb arising from the decay of anthropogenic ²²⁶Ra. Therefore, ²¹⁰Pb is not useful in determining sediment accumulation rates at depth intervals containing anthropogenic ²²⁶Ra.

In light of these complications, supported ²¹⁰Pb activities are estimated from the average of ²²⁶Ra in the upper 0–30 cm and bottom 90–120 cm sediment depths that are assumed not to contain significant anthropogenic ²²⁶Ra. This approach is justified by the close agreement of the ²²⁶Ra activity in these intervals, 1.7 ×10⁻² Bq/g(1.04 dpm/g), with ²²⁶Ra measured elsewhere in the bay. Excess ²¹⁰Pb was then estimated by subtracting the average ²²⁶Ra from the measured ²¹⁰Pb activities. Data are only available from three depth intervals in the sediment over depths from 0 to 30 cm below the surface as shown in Fig. 5. Excess ²¹⁰Pb decays exponentially and was fitted to the simple sedimentation model

$${}^{210}\mathrm{Pb}_{\text{excess}}={}^{210}\mathrm{Pb}_{\text{surface}}\exp (-\lambda z∕s)$$ (1)

where ²¹⁰Pb_surface=3.0×10⁻²±0.2×10⁻² Bq/g(1.8±0.1 dpm/g)=surface excess ²¹⁰Pb; λ=decay rate constant for ²¹⁰Pb (3.09 ×10⁻² year⁻¹), z=depth; and s=sedimentation rate, determined to be 1.37±0.15 cm/year. This ²¹⁰Pb-derived sedimentation rate is greater than the ¹³⁷Cs-derived rate of 0.8±0.2 cm/year.

Fig. 5.

Excess ²¹⁰Pb activity versus depth in Seaplane Lagoon sediment

One method for evaluating the quality of the ²¹⁰Pb data is to compare the flux of excess ²¹⁰Pb to the sediment surface with other studies in the region. The excess ²¹⁰Pb flux is determined by multiplying the surface excess ²¹⁰Pb by the sedimentation rate and the bulk density. At the Seaplane Lagoon the excess ²¹⁰Pb flux is 160 Bq/m²/year (1.0 dpm/cm²/year). This flux is five times higher than the local atmospheric fallout of 33 Bq/m²/year (0.2 dpm/cm²/year) (Fuller and Hammond 1983). Sediment focusing is commonly observed in estuarine sediments (Ivanovich and Harmon 1992), and a sediment ²¹⁰Pb flux 15 times the atmospheric flux has been measured in Richardson Bay, approximately 20 km from the Seaplane Lagoon (Fuller et al. 1999). Within the larger San Francisco Bay system, in situ decay of ²³⁸U in the water column enhanced by exchange between bay waters and Pacific Ocean waters and the erosion within the drainage basin of soils containing excess ²¹⁰Pb have been used to explain the elevated ²¹⁰Pb flux results (Fuller 1982; Fuller et al. 1999). The Seaplane Lagoon is too shallow for in situ decay of ²³⁸U in the water column to be the dominant source of the additional ²¹⁰Pb component. Input of additional ²¹⁰Pb in sediment from the wastewater and storm water outfalls is the most likely source of the nonatmospheric component. The calculated surface ²¹⁰Pb activity for the core is significantly lower than the baywide annual average ²¹⁰Pb activity in suspended sediment, 4.5 ×10⁻² Bq/g(2.7 dpm/g) (Fuller 1982), suggesting that wastewater and storm water outfalls may be the primary source of nonatmospheric excess ²¹⁰Pb activity in lagoon sediments.

Two additional time markers can be used to determine sedimentation rates below sediment depths of 28 cm (1) The Seaplane Lagoon was dredged in 1935 into the Merritt Sand formation. In the sediment cores collected, the top of the Merritt Sand was located at approximately 90 cm depth, and (2) Significant radioactive fallout from atmospheric testing of nuclear weapons began in 1952, so the first appearance of ¹³⁷Cs at 53 cm depth can be used as a chronological marker as well. Both of these time markers below 28 cm indicate a sedimentation rate of approximately 2.2 cm/year between depths of 90 and 28 cm below the sediment-water interface.

Discussion

The ²¹⁰Pb and the ¹³⁷Cs methods for measuring sedimentation rates had a discrepancy in the results, and some justification is required in selecting one result over the other. Differences using these two methods are often the result of either bioturbation of the sediment or loss of the core top during sampling. Bioturbation would mix the greater surface activity of excess ²¹⁰Pb with deeper sediments containing less ²¹⁰Pb. As a result of bioturbation the calculated sedimentation rate is an upper limit. However, the effect of bioturbation is expected to be minimal since the sedimentation rate is relatively high (Olsen et al. 1981) and the high contaminant concentrations and anoxic conditions may deter biological habitation. If the core top was lost, the ¹³⁷Cs-determined sedimentation rate would be higher than actual but there would be no change in the ²¹⁰Pb sedimentation rate. If the upper 19±8 cm of the sediment profile were lost during sampling, then equal sedimentation rates would result from the ²¹⁰Pb and ¹³⁷Cs methods. There is no physical evidence for sediment loss during sampling, and the ¹³⁷Cs estimate is more reliable than the ²¹⁰Pb estimate due to uncertainty caused by the presence of luminescent-based paints. Nonetheless, if the ²¹⁰Pb sedimentation rate is used for the years since 1963, the contaminant reconstruction changes very little because most of the contaminants were deposited prior to 1963.

The time period of contaminant deposition can be approximated using a sedimentation rate of 0.8 cm/year from the surface of the sediment to 28 cm, and 2.2 cm/year between 28 and 90 cm depth. The Cd, Cr, and TPH deposited within the sedimentary interval of 39–85 cm correspond with the years 1958–1938, where the highest levels at a depth of 62 cm represent 1950. The Cu profile is relatively uniform down to a sediment depth of 85 cm (estimated as 1938) followed by a decline at deeper depths. The Pb signature is uniformly high at 600 mg/kg over the sample depths of 39–73 cm corresponding to 1958–1944. The Zn deposited within the sedimentary interval of 39–62 cm corresponds with the years 1958–1950.

The lower metal concentrations at sediment depths of 0–27 cm appear to reflect the diversion and treatment of industrial wastewater from the lagoon outfalls in about 1974. The anthropogenic ²²⁶Ra peak corresponds to about 1944±4, which is consistent with the greatest industrial activity occurring during and after World War II. To demonstrate the degree to which these multiple lines of evidence are consistent, the sedimentary time markers, diversion of industrial wastewater, and peak of ²²⁶Ra are all plotted versus depth in Fig. 6.

Fig. 6.

Sedimentary time markers in sediment from Seaplane Lagoon

Distinguishing the contribution of local sources of contamination from regional sources is often done by comparison with background locations. In some locations, such as the San Francisco Bay, sediment contamination is widespread and has fluctuated with altered industrial activities in the region. This precludes identification of a background level for comparison purposes. Therefore quantifying the contribution of site-specific sources is problematic.

In the Seaplane Lagoon one of the wastewater effluents contained a tracer, ²²⁶Ra, which is most likely unique to Naval Air Station operations. The anthropogenic contribution of ²²⁶Ra as a function of depth in the sediments is determined by using the ²²⁸Ra activity at that depth and the natural activity ratio of ²²⁸Ra/²²⁶Ra to estimate natural ²²⁶Ra levels. Total ²²⁶Ra activity at a given depth minus the calculated natural ²²⁶Ra activity provides an estimate of the anthropogenic contribution. The uniformity of the ²²⁸Ra activity with depth and the reproducibility of the ²²⁶Ra/²²⁸Ra ratio at depths where anthropogenic ²²⁶Ra is not evident both support this approach. Correlation of a contaminant with anthropogenic ²²⁶Ra suggests that the contaminant and ²²⁶Ra had similar release histories. Fig. 7 tests the correlation of the metals and TPH concentrations with the anthropogenic ²²⁶Ra activity. Total petroleum hydrocarbon, Cd, and Cr concentrations correlated with anthropogenic ²²⁶Ra activity and are likely from the same wastewater source or reflect similar levels of industrial activity at the base. Anthropogenic ²²⁶Ra activity and Pb, Zn, and Cu concentrations are not correlated. Although Pb has a distinct peak in the sediment, the lack of correlation with anthropogenic ²²⁶Ra may result from a contaminant source difference such as atmospheric deposition before leaded fuels were phased out for U.S. automobiles or spillage of leaded fuels from refueling of aircraft. The Zn data fluctuate in concentration with depth, but do not have a single distinct peak that correlates with anthropogenic ²²⁶Ra. Finally, no correlation of anthropogenic ²²⁶Ra with Cu should be expected since the Cu concentrations are relatively constant over the depth interval 0–90 cm where significant variations in ²²⁶Ra were observed.

Fig. 7.

Correlation of calculated anthropogenic radium-226 activity with contaminants in Seaplane Lagoon sediment

The engineering profession needs adequate characterization of contaminated sediments prior to undertaking remedial activities. The sediment characterization methodology usually adopted by the maintenance dredging community is based on the collection of sediment cores and homogenizing them over intervals of meters since sediment removal operations will mix sediments over those depths. In contrast, the sediment chronologists are interested in the details of sediment accumulation and strive for sediment analysis over centimeter-depth intervals. Neither approach is practical for contaminated sediment sites that have spatial as well as depth variations. The use of 12-cm intervals in this study permitted determination of sediment accumulation rates and resolved contaminant releases over time intervals of decades. In addition, the ease of gamma counting minimizes sample handling and provides data on radionuclides for age dating, and in this case, identification of an anthropogenic source of ²²⁶Ra. This depth interval resolution was sufficient to assess the time periods when the majority of the contaminants were released into the marina. The approach could be easily extrapolated from a single location to a field site such as the whole marina without excessive analytical costs or modeling efforts.

Conclusion

The composition, quantity, and vertical distribution of contaminants were characterized for only one location in an estuarine marina. Multiple lines of evidence arrived at average sedimentation rates and determined that the greatest contaminant loading was during or soon after World War II. These contaminants have been buried and there is no evidence for their presence in near-surface sediment. This one location is not representative of the entire marina since substantial horizontal heterogeneity was evident from earlier sampling efforts. The ²²⁶Ra in the sediments is specific to discharges from the Naval Air Station and can provides a unique opportunity to delineate the horizontal heterogeneity in contaminant distribution for a site with multiple sources of contamination. The data suggests that ²²⁶Ra measurements could be used as a surrogate for levels of some other industrial wastewater contaminants with a similar release history. Given the ease of ²²⁶Ra analysis, much greater spatial and depth distributions could be determined if warranted. These data suggest that vertical resolution of sediment contamination is essential in understanding historical discharges and current exposures. There is always a trade-off in the value of additional information and the cost of data acquisition that must be addressed during each investigation. While this site has unique features, taking advantage of those features has been beneficial.

Notation

The following symbols are used in this paper:

A_norm

normalized radionuclide activity

A_(z)

radionuclide activity at depth z

A_⁴⁰K(avg)

average ⁴⁰K activity

A_⁴⁰K(z)

⁴⁰K activity at depth z

²¹⁰Pb_excess

excess ²¹⁰Pb at depth z

²¹⁰Pb_surface

excess ²¹⁰Pb at sediment-water interface

s

linear sedimentation rate (cm/year)

z

depth (cm)

λ

²¹⁰Pb decay constant (year⁻¹).