- Email: [email protected]
Assessment of metal concentrations in atmospheric particles from Burnaby Lake, British Columbia, Canada
Atmospheric Environment 35 (2001) 5223–5233
Assessment of metal concentrations in atmospheric particles from Burnaby Lake, British Columbia, Canada R. Brewer*, W. Belzer Aquatic and Atmospheric Sciences Division, Environment Canada, Suite 700-1200 W. 73rd Avenue, Vancouver, British Columbia, Canada V6P 6H9 Received 15 September 2000; received in revised form 26 June 2001; accepted 29 June 2001
Abstract Trace metals were assessed in atmospheric particulates at Burnaby Lake, in the greater Vancouver area of British Columbia to assess concentrations, particle size distributions and deposition rates to an urban watershed. Week-long samples were collected over a period of 18 weeks in 1995 using a 13 stage low pressure impactor (LPI). Samples were analysed using inductively coupled plasma atomic emission spectroscopy (ICP). Aluminum, boron, calcium, iron, magnesium, manganese, sodium and strontium had a similar time series pattern and particle size distribution. For these metals, maximum concentrations occurred during weeks of low precipitation and exhibited a large peak in mid June. Their particle size distribution was mostly dominated by a large peak between 1.7–18.4 mm with a secondary peak at o0.08 mm. Metal concentrations were generally one to three orders of magnitude higher than those measured in a rural location 100 km away from Burnaby Lake but similar to those measured in urban Taipei, Taiwan. Concentrations of the highly toxic metals, arsenic, cadmium and lead were within current air quality guidelines, however boron exceeded the Ontario Ministry of Environment ambient air quality standard in two of the 16 samples. Deposition velocities ranged between 0.22 and 13 cm s1 with the largest values corresponding to the coarse particle mode. Mean deposition rates ranged between 4.0 mg m2 d1 and 650 mg m2 d1. Depending on the metal, yearly loadings to the watershed ranged from 90 kg to several thousand tonnes. Calcium, aluminum, boron and magnesium had the highest metal loadings to the watershed. Manganese also had relatively high loadings, a reflection of the high traffic density in the area. The relatively high metal deposition rates indicate that metal contribution from atmospheric sources may represent a significant portion of the total metal load to the Burnaby Lake watershed. Crown Copyright r 2001 Published by Elsevier Science Ltd. All rights reserved. Keywords: Metals; Deposition rates; Deposition velocities; Particle impactor; Size fractions
1. Introduction Atmospheric particles include all airborne solid and liquid substances. Particles are classified as primary (emitted directly into the atmosphere) or secondary (formed in the atmosphere through chemical and physical transformations). Nucleation mode particles (o0.1 mm in diameter) are formed *Corresponding author. Tel.: +1-604-664-4070; fax: +1604-664-9126. E-mail address: [email protected] (R. Brewer).
primarily from the condensation of hot vapours during high temperature combustion processes and from the nucleation of atmospheric gas species to form new particles. Particles in the accumulation mode (0.1–2.0 mm) result from the coagulation of particles in the nucleation mode and from the condensation of vapours onto existing particles. Particles larger than 2.0 mm (sedimentation or coarse mode) are typically associated with mechanical processes such as wind erosion, marine aerosols and grinding operations that generate windblown soil, sea salt spray and dust.
1352-2310/01/$ - see front matter Crown Copyright r 2001 Published by Elsevier Science Ltd. All rights reserved. PII: S 1 3 5 2 - 2 3 1 0 ( 0 1 ) 0 0 3 4 3 - 0
5224
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
Atmospheric particles are of concern due to their impacts on visibility, human health, plants, aquatic life and materials. Visibility impairment is caused by the scattering and absorption of light by atmospheric aerosols, particularly those of o2.5 mm (National Research Council, 1993). Human impacts are largely health-related and generally associated with inhalation of particulate matter of o10 mm in diameter. Both acute and chronic effects have been linked to exposure of inhalable particles with a demonstrated decrease in lung function, increased respiratory disease symptoms and increased mortality (CEPA/FPAC, 1998). Respirable particles (o2.5 mm in diameter) can reach the lung alveoli and therefore are especially important in the absorption of metals and other toxic contaminants. Detrimental effects on vegetation include reduced growth and productivity due to interference with photosynthesis and phytotoxic impacts. Aquatic life can be affected through ingestion and adsorption of particulate matter deposited or washed into water bodies. Buildings and materials are negatively affected through increased rates of corrosion, erosion, soiling and discoloration. The present study was conducted under Environment Canada’s Fraser River Action Plan as part of a larger ecosystem study designed to assess the atmospheric
contribution of contaminants of concern to the Still Creek-Burnaby Lake-Brunette River ecosystem. In this study metals were measured in atmospheric aerosols at Burnaby Lake, located in a highly urbanised and industrialised area within the municipalities of Burnaby, Vancouver and New Westminster, British Columbia (Fig. 1). Anthropogenic sources of trace metals to the watershed are typically dominated by vehicular and industrial emissions. The objectives of this study were (1) to characterise the concentration and particle size distribution of metals in ambient air in the Burnaby Lake area and (2) to quantify atmospheric deposition of metals to the watershed.
2. Site description Burnaby Lake is located between two ridges that run laterally for several kilometers in an approximately east– west direction on the north and south sides of the lake. Because of the topography, the natural flow of winds though the lake basin is generally from the east or the west. The watershed drains 6060 ha of land divided into four predominant land uses: 42% residential, 31% open space and forested, 15% commercial and institutional
Fig. 1. Map of sampling site at Burnaby Lake and surrounding area.
5225
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
and 5.5% industrial (Hall and Anderson, 1988). Approximately 15% of the total land area is taken up by highways, streets and alleys. This includes two major highways, which follow the lower elevation contours around both sides of Burnaby Lake and form the major east–west corridor between Vancouver and the Fraser Valley. The commercial and institutional areas are scattered throughout the basin, while the industrial areas (mechanical, shipping, transportation and manufacturing) are concentrated along Still Creek to the west of Burnaby Lake (Hall and Anderson, 1988). The north, south and southwest sides are surrounded by residential areas. Highway 7 runs along the north side of the lake and Highway 1 along the south side. A major intersection connects these two highways to the east (Fig. 1). Atmospheric sampling equipment was installed on the roof of the Rowing Club Pavilion Building on the south side of Burnaby Lake at a height of 4.9 m above water. The building was situated over the lake on pilings and was away from any local sources of pollution. The site was chosen for its proximity to traffic corridors, as well as its distance from any particular point source. Because of the geography of the site which rendered it free of physical obstructions to wind flow, samples were expected to be representative of the area.
3. Field and laboratory methods 3.1. Meteorological measurements A meteorological tower was used for recording physical variables such as wind direction, wind speed,
temperature, barometric pressure and relative humidity. Wind speed and direction were measured using an RM Young anemometer mounted at a height of 7.9 m above the lake and 3 above sampling height. The anemometer was placed at a distance 10 times the tallest object in the vicinity of the site. Temperature was measured using an RM Young thermistor. Barometric pressure was measured with a Setra Systems aneroid barometer. Relative humidity was measured with a RM Young hygrometer. Rainfall amount was measured using an Environment Canada standard rain gauge. Meteorological data were captured by a CR-10 data logger from Campbell Scientific. 3.2. Low pressure impactor measurements Atmospheric particulate matter was collected with a 13 stage low pressure cascade impactor (LPI) Model # 20-930 capable of separating atmospheric particles into unique size fractions corresponding to the following midpoint diameters (mm): >35, 28.4, 18.4, 13.1, 8.6, 5.0, 2.7, 1.7, 1.2, 0.71, 0.38, 0.17, 0.10 and o0.08. Sampling procedures were those established by the manufacturer (Graseby Andersen, Smyrna, Georgia). Glass fiber filters were used as substrates for the impaction plates and backup filter. Substrates were sprayed according to manufacturer’s specifications with silicone spray (Dow Corning 316) to diminish particle bounce and improve capture. Samples were collected for 18 weeks between 11 April and 17 October 1995 (Table 1). Individual samples were collected over a period of seven days with the exception of the August 29 sample which was collected over 14 days. The sampling volume ranged
Table 1 Sample collection information for LPI samples Week
Date on
Time on
Date off
Time off
Time elapsed (min)
Flow rate (l min1)
Volume of air sampled (m3)
1 2 3 4 5 6 7 8 9 10 11,12 13 14 15 16 17
11-Apr-95 25-Apr-95 16-May-95 30-May-95 13-Jun-95 04-Jul-95 11-Jul-95 25-Jul-95 08-Aug-95 22-Aug-95 29-Aug-95 12-Sep-95 19-Sep-95 26-Sep-95 03-Oct-95 10-Oct-95
09:10 08:45 09:00 07:40 10:00 08:50 08:55 09:15 09:00 08:00 08:00 09:30 08:30 08:50 10.30 09:05
18-Apr-95 02-May-95 23-May-95 06-Jun-95 20-Jun-95 11-Jul-95 18-Jul-95 01-Aug-95 15-Aug-95 29-Aug-95 12-Sep-95 19-Sep-95 26-Sep-95 03-Oct-95 10-Oct-95 17-Oct-95
08:30 09:30 09:00 07:30 07:30 08:10 07:30 08:00 09:00 07:55 09:00 08:00 08:10 10:00 08:30 08:00
10,040 10,125 10,060 10,070 9930 10,040 10,055 10,005 10,060 10,075 20,220 9990 10,060 10,150 9960 10,015
3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0
30.1 30.4 30.0 30.2 29.8 30.1 30.2 30.0 30.0 30.2 60.7 30.0 30.2 30.5 29.9 30.0
5226
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
from 29.8–60.7 m3. The flow rate (3.0 l min1) was calculated from the pressure drop across the critical orifice using a calibration graph provided by the manufacturer.
Deposition velocities for particles >5 mm were calculated assuming negligible gravitational effects using the following formula:
3.3. Field quality assurance/quality control methods
where
To ensure that the meteorological instruments were performing to standard, they were evaluated yearly using manufacturer’s procedures to ensure that they performed accurately and within the manufacturer’s specifications. Field blanks were used to assess sample handling by subjecting them to the same field handling and laboratory analysis as the samples. Sample handling quality assurance protocols were established and documented prior to starting the sampling program to ensure consistent procedures and reduce variability in the data. Personnel were required to wear polyethylene gloves during sample handling.
* Vdð>5 mmÞ ¼ EA uavg ðm s1 Þ;
ð2Þ
EA ¼ the particle effective inertial coefficient ¼ 1:12e30:36=da ;
ð3Þ
where da is the aerodynamic particle diameter * ¼ the mean friction velocity uavg
¼ kuavg ln½Z0 =ðZ dÞ ðm s1 Þ;
ð4Þ
where k=Von Karman’s constant (0.4), Z0 =surface roughness (0.25 m) after Noll et al. (1985) for a mixed institutional, commercial and residential area, Z=height of the wind speed measurement (m), d=height of the sampler above ground (m).
3.4. Laboratory methods Thirty two metals were measured in LPI particulate samples (Table 2). Samples were analysed at ZENON Analytical Services, Vancouver, BC, using inductively coupled plasma atomic emission spectroscopy (ICP) (BCMELP, 1983). The QA/QC component of the analytical analysis consisted of substrate and reagent blanks. Blanks were used to assess reagents, methods, standards, instruments and calibrations.
4. Data processing and analysis 4.1. Concentrations Raw data were blank corrected based on averaged field blank values. Values below the minimum detection limit were set to zero. This was chosen in favor of the 12 detection limit option in order to minimise overestimation error, as the detection limit was high for a number of elements. Concentrations were calculated from blank corrected data by dividing sample weights by the sample volume as reported in Table 1. 4.2. Deposition velocities Deposition velocities (Vd ) for particles p5 mm were calculated after Slinn and Slinn (1980), while those for particles >5 mm were calculated after Noll and Fang (1989). For particles p5 mm, the deposition velocity was calculated using the following formula: Vdðo5 mmÞ ¼ CD uavgðhÞ ðm s1 Þ;
ð1Þ 3
where CD =the drag coefficient (1.3 10 ), uavgðhÞ =the mean wind velocity at height h (m s1).
Table 2 Target analytes for the analysis of trace metals and their detection limits (mg m3) Analyte
Detection limit
Ag Al As B Ba Be Bi Ca Cd Co Cr Cu Fe K Mg Mn Mo Na Ni P Pb S Sb Se Sn Sr Te Ti Tl V Zn Zr
0.00027 0.17 0.0013 0.10 0.0025 0.00050 0.066 1.3 0.15 0.0075 0.050 1.5 1.2 1.0 0.050 0.025 0.010 1.0 0.050 0.10 1.2 0.25 0.35 0.0013 0.075 0.43 0.050 0.17 0.075 0.012 0.43 0.015
5227
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
log transformed and standardised to standard deviation. Metals with >5 non-detectable values were not included in the PCA analysis. Principal components were extracted using a correlation matrix.
4.3. Deposition rates Deposition rates were calculated from the product of the deposition velocity and metal concentration in each particle size fraction. Total deposition (Dt ) was calculated as the sum of deposition rates for each particle size interval. The equation consisted of a combination of the Slinn and Slinn (1980) and Noll and Fang (1989) equations for particles of diameter p5 and >5 mm, respectively X X Dt ¼ ½ð1:3103 Þuavg Cda þ ðda o5 mmÞ
5. Results and discussion 5.1. Meteorological data Meteorological data collected during the study period are presented in Table 3. Results represent average weekly values. Some meteorological data was lost due to data logger overload or adverse weather conditions. The mean temperature ranged from a low of 7.91C to a high of 19.81C. The mean relative humidity ranged from 64.6–92.5% with lower values typical of the summer months. Precipitation varied throughout the study period, ranging from 0.4–89.8 mm with considerably greater amounts of precipitation falling during the fall months. Mean atmospheric pressure values ranged from 1009–1018 mbar. Wind speeds ranged from 4.8– 6.8 km h1. Winds were predominantly from the east during the study period.
ðda >5 mmÞ
h i 2 * ÞCd ð1:12e30:36=da uavg sÞ; a ðng m
ð5Þ
where da =the aerodynamic particle diameter, Cda =the contaminant concentration for particles of aerodynamic diameter ‘‘a’’ (ng m3), uavg¼ mean wind speed (m s1), u avg =mean friction velocity (m s1) 4.4. Principal components analysis Principal component analysis (PCA) was used for source interpretation. Source types were identified by associating the statistically significant component loadings with elements related to specific source types. Data were arranged in a matrix such that each column contained either the concentration of one chemical species or a physical parameter and each row a single weekly sample. Prior to running the analysis, data were
5.2. Metals data Of the 32 metals measured in atmospheric aerosols at Burnaby Lake, 14 were assessed for particle size
Table 3 Meteorological data collected during the sampling perioda Week
Start
Stop
Wind directionb
Wind speed (km h1)c
Temperature (1C)
Pressure (mbar)
RHd (%)
Precipitation (mm)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
11-Apr-95 25-Apr-95 16-May-95 30-May-95 13-Jun-95 04-Jul-95 11-Jul-95 25-Jul-95 08-Aug-95 22-Aug-95 29-Aug-95 12-Sep-95 19-Sep-95 26-Sep-95 03-Oct-95 10-Oct-95
17-Apr-95 28-Apr-95 22-May-95 05-Jun-95 19-Jun-95 10-Jul-95 17-Jul-95 31-Jul-95 13-Aug-95 28-Aug-95 04-Sep-95 18-Sep-95 25-Sep-95 02-Oct-95 09-Oct-95 16-Oct-95
S, SW SW, SE SW SE and S SE, E, W and SW SE SE, NW SW and SE E S E, NE NE NE E and NE E SE
6.2 5.5 5.9 5.9 6.2 5.3 6.7 6.8 6.7 5.1 n/ae n/a n/a n/a 6.1 4.8
7.9 15.1 14.4 16.8 14.9 17.9 19.8 17.4 15.6 15.8 17.0 18.2 15.7 13.2 11.5 11.3
1013 1009 1016 1013 1010 1014 1018 1015 1014 1014 1014 1015 1015 1012 1015 1016
78.1 64.6 69.3 73.9 83.3 78.9 73.1 75.9 85.6 80.1 80.5 82.6 76.0 90.0 90.9 92.5
24.0 0.0 0.4 8.8 9.2 19.6 0.0 51.0 11.2 7.2 9.4 1.6 4.4 33.6 89.8 76.6
a
Values represent means for each sampling period. Prevailing wind direction. c Average wind speed derived from 5 min intervals within the sampling period. d Relative humidity. e n/a denotes data not available due to equipment failure. b
5228
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
distribution, 12 were assessed for total concentrations for the sum of particle size fractions and six were not detected (Table 4). The metals assessed for total Table 4 Summary of assessment of metals in LPI samples from Burnaby Lake Assessed for particle size distributiona
Assessed for total concentrationsb
Not detected
Al As B Ba Ca Fe Mg Mn Na S Sr Ti V Zr
Ag Be Cd Co Cr K Mo P Pb Se Te Zn
Bi Cu Ni Sb Sn Tl
a b
Size fractions quantifiable. Size fractions not quantifiable.
concentrations could not be assessed for individual size fractions due to a high frequency of non-detectable values in the size fractions. Table 5 presents a statistical summary of metal concentrations for the sum of particle fractions. PCA analysis indicated a strong association between aluminum, barium, boron, calcium, iron, magnesium, manganese, sodium, strontium, and vanadium (Fig. 2). Further examination confirmed that most of these metals had a similar time series and particle size distribution. Maximum concentrations occurred during weeks of low precipitation with a large peak for the week including 13 June followed by multiple smaller peaks in the late summer-early fall period (Fig. 3). This is consistent with a westward transport of marine aerosols in the region and elevated concentrations of geological material (Chow and Watson, 1998). Concentration maxima for titanium and vanadium occurred in the late summer and early fall (not shown). While titanium is commonly associated with soil dusts, vanadium is often present in emissions from residual oil combustion (Chow and Watson, 1998). The particle size distribution for this group of metals was mostly dominated by a large peak between 1.7 and 18.4 mm, with a smaller secondary peak at o0.08 mm (Figs. 3 and 4). For the
Table 5 Statistical summary of trace metal concentrations (mg m3) for the sum of particle size fractions (n ¼ 16) Analyte
Detection limit (DL)
Mean
S.D.
Minimum
Maximum
Fraction of data with values 4DL (%)
Ag Al As B Ba Be Ca Cd Co Cr Fe K Mg Mn Mo Na P Pb S Se Sr Te Ti V Zn Zr
0.00027 0.17 0.0013 0.10 0.0025 0.00050 1.3 0.15 0.0075 0.050 1.2 1.0 0.050 0.025 0.010 1.0 0.10 1.2 0.25 0.0013 0.43 0.050 0.17 0.012 0.43 0.015
0.0014 200 0.023 55 0.24 0.0024 480 0.00039 0.0050 0.012 4.8 4.0 48 0.17 0.0018 26 0.45 0.049 6.8 0.0022 7.3 0.010 1.7 0.064 0.084 0.20
0.0038 190 0.068 50 0.30 0.0053 450 0.0012 0.0092 0.032 4.8 2.4 45 0.19 0.0052 21 0.22 0.043 4.3 0.0046 6.3 0.030 0.92 0.069 0.23 0.13
0.002 13 0.0012 3.3 0.0052 0.0004 41 0.0018 0.008 0.025 1.3 1.1 3.6 0.032 0.011 5.0 0.27 0.010 1.6 0.0015 1.2 0.053 0.39 0.032 0.59 0.051
0.015 760 0.28 180 1.1 0.022 1800 0.0045 0.030 0.12 18 8.8 180 0.77 0.018 91 0.89 0.14 18 0.014 26 0.11 3.8 0.23 0.76 0.46
31 100 69 100 100 63 100 13 31 19 81 94 100 88 13 100 100 88 100 25 100 13 100 63 13 100
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
Fig. 2. PCA loadings plot of sample metal concentrations for the sum of particle fractions (n ¼ 16). Factors 1 and 2 explained 53% and 13% of the variability, respectively.
5229
coarse particle fraction the source is likely soil dust and sea salt, while metals in the submicron particle range are likely from various anthropogenic sources. Aluminum, sodium and manganese have been shown to be present in fine particles from diesel and gasoline combustion (Kleeman et al., 2000), calcium and magnesium in lube oil combustion emissions, and submicron particles of iron, magnesium and manganese are emitted from engine wear or fuel components (Graham, unpublished). Manganese is commonly used in the anti-knock agent methylcyclopenpentadienyl manganese tricarbonyl (MMT) in some types of commercial gasolines (Chow and Watson, 1998). Strontium is an alkaline earth metal normally associated with sulphate or carbonate, and boron occurs naturally in soils and as a bi-product of glass manufacturing (Merck Index, 1989). The grouping of the remaining metals in the PCA plot suggested different sources. For example, lead was most closely associated with beryllium (Fig. 2), possibly
Fig. 3. Typical concentration time series for: Al (shown here), B, Ca, Fe, Mg, Mn, Na and Sr for the sum of particle size fractions (solid line, mg m3) and precipitation (dashed line, mm) during sampling period.
Fig. 4. Typical particle size distribution for: Al (shown here), Ba, B, Ca, Fe, Mg, Mn, Na, Sr, Ti and V. Concentrations are expressed as means (mg m37SE), n ¼ 16:
5230
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
reflecting electrical manufacturing sources. Phosphorus and potassium were not associated with any other variables, pointing to unique sources such as agricultural fertilisers for the former and vegetative burning for the latter (Chow and Watson, 1998). Sulphur was associated with wind direction (which was generally from the south, southwest during the two weeks with peak sulphur concentrations (Fig. 5)) indicating marine sources and or inputs from the oil refineries south of the Canada–US border. Sulphur concentration maxima occurred during the spring period (between 11 April and 23 May) and corresponded to a period of low precipitation (Fig. 5). Furthermore, mean annual sulphur concentrations exhibited a bimodal distribution corresponding to coarse and fine particles, respectively (Fig. 6). The fine particle fraction is likely dominated by ammonium sulphate aerosols (Pryor et al., 1997), while the coarse fraction is probably largely composed of sea salt sulphate (Savoy and Prospero, 1982). Seasonally,
sulphur concentrations were highest in the fine particle range during the spring (Fig. 7), suggesting that marine sources, likely in the form of dimethyl sulphide precursors, may play an important role (Kettle et al., 1999). Metal concentrations were generally one to three orders of magnitude higher than those measured 100 km away in rural Abbotsford in 1996 but similar to those measured in urban Taipei, Taiwan (Li et al., 1993). Exceptions were found for aluminum and calcium which were in higher concentrations at Burnaby Lake. Concentrations of the highly toxic metals, arsenic, cadmium and lead, were within current air quality guidelines. Lead was interesting in that it’s time series pattern was similar to that of the crustal elements, suggesting an association with soil dusts. Although lead has been banned from use as a gasoline additive in Canada for over a decade, residual levels are likely to persist in urban particulate matter for many decades to come.
Fig. 5. Sulphur concentration time series for the sum of particle size fractions (solid line, mg m3) and precipitation (dashed line, mm) during the sampling period.
Fig. 6. Particle size distribution of sulphur. Concentrations are expressed as means (mg m37SE), n ¼ 16:
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
Boron exceeded the Ontario Ministry of Environment ambient air quality standard (OME (Ontario Ministry of Environment), 1999) in two of the 16 samples. Deposition velocities ranged between 0.22 and 13 cm s1 and increased with particle diameter (Fig. 8). It should be noted that the errors associated with calculation of deposition velocities are greater for larger width particle fractions, as particle fraction width increases with mean diameter size. This results in larger errors of the estimated deposition rate for the larger particle size fractions. Total deposition rates varied for each metal with mean values ranging between
5231
4.0 mg m2 d1 and 650 mg m2 d1(Table 6). Deposition rates generally reflected concentration patterns with the highest deposition rates occurring for the week including 13 June. Depending on the metal, yearly loadings to the watershed ranged from 90 kg to several thousand tonnes (Table 7). Calcium, aluminum, boron and magnesium had the highest metal loadings to the watershed. Manganese also had relatively high loadings and was reported to exceed federal and provincial water quality guidelines in creeks and rivers within the Burnaby Lake watershed (Sekela et al., 1998), which may support the atmospheric pathway as an important source of this
Fig. 7. Sulphur concentration for the different size fractions during the spring (n ¼ 5), summer (n ¼ 6) and fall (n ¼ 5). Spring: 11 April–20 June ; Summer: 4 July–12 September; Fall: 12 September–17 October.
Fig. 8. Calculated deposition velocities (cm s1) as a function of particle size.
5232
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
Table 6 Statistical summary of trace metal deposition rates (mg m-2 d1) for the sum of particle size fractions (11 April–10 October 1995) (n ¼ 11) Analytea
Mean
S.D.
Minimum
Maximum
Al As B Ba Be Ca Fe K Mg Mn Na P Pb S Sr Ti V Zr
270,000 100 69,000 260 4.0 650,000 6000 11,000 64,000 350 44,000 920 78 23,000 10,000 2000 61 330
360,000 330 87,000 420 10 870,000 7800 9000 84,000 500 47,000 640 120 12,000 13,000 1400 72 230
10,000 2.1 2900 0.86 0.29 21,000 620 420 2600 6.6 840 270 1.0 8000 850 84 6.6 12
1,300,000 1100 300,000 1100 34 3,100,000 27,000 26,000 300,000 1500 170,000 2300 380 47,000 45,000 4300 180 600
a
Deposition values were not calculated for Ag, Cd, Co, Cr, Mo, Se, Te and Zn due to >50% of the samples being below the detection limit.
Table 7 Estimate of yearly metal deposition rates to entire Burnaby Lake watershed
contribute significantly to the total metal loading to urban watersheds (Chow and Watson, 1998), resulting in an underestimation of the total metal deposition to the Burnaby Lake watershed. Long range transport and deposition likely accounts for a portion of the total atmospheric load to the area, however the magnitude of this contribution is unknown.
6. Conclusion The air pathway appears to be an important source of metals to the Burnaby Lake area. The relatively high metal concentrations measured at the site are a reflection of the highly urbanised location of the lake, receiving inputs from a multitude of natural and anthropogenic sources. The exceedence of the Ontario Ministry of Environment ambient air quality standard for boron in several samples suggests that concentrations may occasionally pose a risk to human health. The relatively high metal deposition rates to the area indicates that the metal contribution from atmospheric sources may represent a significant portion of the total metal load to the Burnaby Lake watershed, while the reported exceedence of water quality guidelines and criteria for several metals in local rivers suggests that aquatic life may be at risk in the more urbanised reaches.
Analyte
Estimated deposition ratea (ton yr1)
References
Al As B Ba Be Ca Fe K Mg Mn Na P Pb S Sr Ti V Zr
5900 2.3 1527 5.7 0.090 14,000 130 240 1400 7.7 980 20 1.7 500 230 45 1.3 7.3
BCMELP (British Columbia Ministry of the Environment, Lands and Parks), 1983. A laboratory manual for the chemical analysis of ambient air emissions, precipitation, soil and vegetation. In: Neil McQuacker (Ed.), Waste Management Branch. Data Standards Group, Vancouver, BC. CEPA/FPAC (Canadian Environmental Protection Act/ Federal Provincial Advisory Committee), 1998. National Ambient Air Quality Objectives for Particulate Matter. Part 1: Science Assessment Document. A report by the CEPA/FPAC Working Group on Air Quality Objectives and Guidelines, ISBN 0-662-63486-1, 19pp. Chow, J. C., Watson, J. G., 1998. Western Washington 1996–97 PM 2.5 Source Apportionment Study. Worldwide Environmental Corporation, 8101 Meadow Vista Dr., Reno, Nevada, 89511. Prepared for Puget Sound Air Pollution Control Agency, 110 Union Street, Suite 500, Seattle, Washington, 98101. Graham, L., Gaseous and Particulate Matter Emissions from In-Use Light Duty Gasoline Motor Vehicles. Environment Canada, Environmental Technology Advancement Directorate, Environmental Technology Centre, Ottawa, Ontario, K1A 0H3, EDM Report #99-67, 90pp (unpublished). Hall, K.J., Anderson, B.C., 1988. The toxicity and chemical composition of stormwater runoff. Canadian Journal of Civil Engineers 15, 98–106.
a
Based on an area of 6060 ha.
metal. It should be noted that due to high background levels in the blanks, loadings shown in Table 7 do not include some of the common elements associated with urban aerosols, such as Si, Cr, Cu and Zn, which can
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233 Kettle, A.J., Andreae, M.O., Amouroux, D., Andreae, T.W., Bates, T.S., Berresheim, H., Bingemer, H., Boniforti, R., Curran, M.A.J., DiTullio, G.R., Helas, G., Jones, G.B., Keller, M.D., Kiene, R.P., Leck, C., Levasseur, M., Malin, G., Maspero, M., Matrai, P., McTaggard, A.R., Mihalopoulos, N., Nguyen, B.C., Novo, A., Putaud, J.P., Rapsomanikis, S., Roberts, G., Schebeske, G., Sharma, S., Simo, R., Staubes, R., Turner, S., Uher, G., 1999. A global database of sea surface dimethylsulphide (DMS) measurements and a procedure to predict sea surface DMS as a function of latitude, longitude, and month. Global Biogeochemical Cycles 13 (2), 399–444. Kleeman, M.J., Schauer, J.J., Cass, G.R., 2000. Size and composition distribution of fine particulate matter emitted from motor vehicles. Environmental Science and Technology 34 (7), 1132–1142. Li, C.-S., Hsu, L.-Y., Chuang, Y.-Y.T., 1993. Elemental profiles of indoor and outdoor particulate matter less than 10 mm (PM 10) and 2.5 mm (PM 2.5) in Taipei. Chemosphere 27 (11), 2143–2154. Merck Index, 1989. The Merck Index. In: Susan Budavari (Ed.), An Encyclopedia of Drugs and Chemicals, eleventh edition. Merck & Co. Inc. Rahway, NJ, USA, 1606 pp + app. National Research Council, 1993. Protecting Visibility in National Parks and Wilderness Areas. National Academy Press, Washington, DC, 446pp.
5233
Noll, K.E., Pontius, A., Frey, R., Gould, M., 1985. Comparison of atmospheric coarse particles at an urban and nonurban site. Atmospheric Environment 19, 1931–1943. Noll, K.E., Fang, K.Y.P., 1989. Development of a dry deposition model for atmospheric coarse particles. Atmospheric Environment 23 (3), 585–594. OME (Ontario Ministry of Environment), 1999. Summary of Point of Impingement Standards, Point of Impingement Guidelines and Ambient Air Quality Criteria (AAQCs). Standards and Development Branch, Ontario Ministry of Environment, 10pp. Pryor, S.C., Barthelmie, R.J., Hoff, R.M., Sakiyama, S., Simpson, R., Steyn, D., 1997. Reveal: characterizing fine aerosols in the Fraser Valley, BC. Atmosphere-Ocean 35 (2), 209–227. Savoy, D.L., Prospero, J.M., 1982. Particle size distribution of nitrate and sulfate in the marine atmosphere. Geophysical Research Letters 9 (10), 1207–1210. Sekela, M., Brewer, R., Tuominen, T., Sylvestre, S., Moyle G., 1998. Effect of a Rainfall Event on Contaminant Levels in the Brunette River Watershed. Environment Canada, Aquatic and Atmospheric Sciences Division, Environmental Conservation Branch, Pacific and Yukon Region. Vancouver, BC, DOE FRAP 1997-36, 35pp. Slinn, S.A., Slinn, W.G., 1980. Predictions for particle deposition on natural waters. Atmospheric Environment 14, 1013–1016.
Assessment of metal concentrations in atmospheric particles from Burnaby Lake, British Columbia, Canada R. Brewer*, W. Belzer Aquatic and Atmospheric Sciences Division, Environment Canada, Suite 700-1200 W. 73rd Avenue, Vancouver, British Columbia, Canada V6P 6H9 Received 15 September 2000; received in revised form 26 June 2001; accepted 29 June 2001
Abstract Trace metals were assessed in atmospheric particulates at Burnaby Lake, in the greater Vancouver area of British Columbia to assess concentrations, particle size distributions and deposition rates to an urban watershed. Week-long samples were collected over a period of 18 weeks in 1995 using a 13 stage low pressure impactor (LPI). Samples were analysed using inductively coupled plasma atomic emission spectroscopy (ICP). Aluminum, boron, calcium, iron, magnesium, manganese, sodium and strontium had a similar time series pattern and particle size distribution. For these metals, maximum concentrations occurred during weeks of low precipitation and exhibited a large peak in mid June. Their particle size distribution was mostly dominated by a large peak between 1.7–18.4 mm with a secondary peak at o0.08 mm. Metal concentrations were generally one to three orders of magnitude higher than those measured in a rural location 100 km away from Burnaby Lake but similar to those measured in urban Taipei, Taiwan. Concentrations of the highly toxic metals, arsenic, cadmium and lead were within current air quality guidelines, however boron exceeded the Ontario Ministry of Environment ambient air quality standard in two of the 16 samples. Deposition velocities ranged between 0.22 and 13 cm s1 with the largest values corresponding to the coarse particle mode. Mean deposition rates ranged between 4.0 mg m2 d1 and 650 mg m2 d1. Depending on the metal, yearly loadings to the watershed ranged from 90 kg to several thousand tonnes. Calcium, aluminum, boron and magnesium had the highest metal loadings to the watershed. Manganese also had relatively high loadings, a reflection of the high traffic density in the area. The relatively high metal deposition rates indicate that metal contribution from atmospheric sources may represent a significant portion of the total metal load to the Burnaby Lake watershed. Crown Copyright r 2001 Published by Elsevier Science Ltd. All rights reserved. Keywords: Metals; Deposition rates; Deposition velocities; Particle impactor; Size fractions
1. Introduction Atmospheric particles include all airborne solid and liquid substances. Particles are classified as primary (emitted directly into the atmosphere) or secondary (formed in the atmosphere through chemical and physical transformations). Nucleation mode particles (o0.1 mm in diameter) are formed *Corresponding author. Tel.: +1-604-664-4070; fax: +1604-664-9126. E-mail address: [email protected] (R. Brewer).
primarily from the condensation of hot vapours during high temperature combustion processes and from the nucleation of atmospheric gas species to form new particles. Particles in the accumulation mode (0.1–2.0 mm) result from the coagulation of particles in the nucleation mode and from the condensation of vapours onto existing particles. Particles larger than 2.0 mm (sedimentation or coarse mode) are typically associated with mechanical processes such as wind erosion, marine aerosols and grinding operations that generate windblown soil, sea salt spray and dust.
1352-2310/01/$ - see front matter Crown Copyright r 2001 Published by Elsevier Science Ltd. All rights reserved. PII: S 1 3 5 2 - 2 3 1 0 ( 0 1 ) 0 0 3 4 3 - 0
5224
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
Atmospheric particles are of concern due to their impacts on visibility, human health, plants, aquatic life and materials. Visibility impairment is caused by the scattering and absorption of light by atmospheric aerosols, particularly those of o2.5 mm (National Research Council, 1993). Human impacts are largely health-related and generally associated with inhalation of particulate matter of o10 mm in diameter. Both acute and chronic effects have been linked to exposure of inhalable particles with a demonstrated decrease in lung function, increased respiratory disease symptoms and increased mortality (CEPA/FPAC, 1998). Respirable particles (o2.5 mm in diameter) can reach the lung alveoli and therefore are especially important in the absorption of metals and other toxic contaminants. Detrimental effects on vegetation include reduced growth and productivity due to interference with photosynthesis and phytotoxic impacts. Aquatic life can be affected through ingestion and adsorption of particulate matter deposited or washed into water bodies. Buildings and materials are negatively affected through increased rates of corrosion, erosion, soiling and discoloration. The present study was conducted under Environment Canada’s Fraser River Action Plan as part of a larger ecosystem study designed to assess the atmospheric
contribution of contaminants of concern to the Still Creek-Burnaby Lake-Brunette River ecosystem. In this study metals were measured in atmospheric aerosols at Burnaby Lake, located in a highly urbanised and industrialised area within the municipalities of Burnaby, Vancouver and New Westminster, British Columbia (Fig. 1). Anthropogenic sources of trace metals to the watershed are typically dominated by vehicular and industrial emissions. The objectives of this study were (1) to characterise the concentration and particle size distribution of metals in ambient air in the Burnaby Lake area and (2) to quantify atmospheric deposition of metals to the watershed.
2. Site description Burnaby Lake is located between two ridges that run laterally for several kilometers in an approximately east– west direction on the north and south sides of the lake. Because of the topography, the natural flow of winds though the lake basin is generally from the east or the west. The watershed drains 6060 ha of land divided into four predominant land uses: 42% residential, 31% open space and forested, 15% commercial and institutional
Fig. 1. Map of sampling site at Burnaby Lake and surrounding area.
5225
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
and 5.5% industrial (Hall and Anderson, 1988). Approximately 15% of the total land area is taken up by highways, streets and alleys. This includes two major highways, which follow the lower elevation contours around both sides of Burnaby Lake and form the major east–west corridor between Vancouver and the Fraser Valley. The commercial and institutional areas are scattered throughout the basin, while the industrial areas (mechanical, shipping, transportation and manufacturing) are concentrated along Still Creek to the west of Burnaby Lake (Hall and Anderson, 1988). The north, south and southwest sides are surrounded by residential areas. Highway 7 runs along the north side of the lake and Highway 1 along the south side. A major intersection connects these two highways to the east (Fig. 1). Atmospheric sampling equipment was installed on the roof of the Rowing Club Pavilion Building on the south side of Burnaby Lake at a height of 4.9 m above water. The building was situated over the lake on pilings and was away from any local sources of pollution. The site was chosen for its proximity to traffic corridors, as well as its distance from any particular point source. Because of the geography of the site which rendered it free of physical obstructions to wind flow, samples were expected to be representative of the area.
3. Field and laboratory methods 3.1. Meteorological measurements A meteorological tower was used for recording physical variables such as wind direction, wind speed,
temperature, barometric pressure and relative humidity. Wind speed and direction were measured using an RM Young anemometer mounted at a height of 7.9 m above the lake and 3 above sampling height. The anemometer was placed at a distance 10 times the tallest object in the vicinity of the site. Temperature was measured using an RM Young thermistor. Barometric pressure was measured with a Setra Systems aneroid barometer. Relative humidity was measured with a RM Young hygrometer. Rainfall amount was measured using an Environment Canada standard rain gauge. Meteorological data were captured by a CR-10 data logger from Campbell Scientific. 3.2. Low pressure impactor measurements Atmospheric particulate matter was collected with a 13 stage low pressure cascade impactor (LPI) Model # 20-930 capable of separating atmospheric particles into unique size fractions corresponding to the following midpoint diameters (mm): >35, 28.4, 18.4, 13.1, 8.6, 5.0, 2.7, 1.7, 1.2, 0.71, 0.38, 0.17, 0.10 and o0.08. Sampling procedures were those established by the manufacturer (Graseby Andersen, Smyrna, Georgia). Glass fiber filters were used as substrates for the impaction plates and backup filter. Substrates were sprayed according to manufacturer’s specifications with silicone spray (Dow Corning 316) to diminish particle bounce and improve capture. Samples were collected for 18 weeks between 11 April and 17 October 1995 (Table 1). Individual samples were collected over a period of seven days with the exception of the August 29 sample which was collected over 14 days. The sampling volume ranged
Table 1 Sample collection information for LPI samples Week
Date on
Time on
Date off
Time off
Time elapsed (min)
Flow rate (l min1)
Volume of air sampled (m3)
1 2 3 4 5 6 7 8 9 10 11,12 13 14 15 16 17
11-Apr-95 25-Apr-95 16-May-95 30-May-95 13-Jun-95 04-Jul-95 11-Jul-95 25-Jul-95 08-Aug-95 22-Aug-95 29-Aug-95 12-Sep-95 19-Sep-95 26-Sep-95 03-Oct-95 10-Oct-95
09:10 08:45 09:00 07:40 10:00 08:50 08:55 09:15 09:00 08:00 08:00 09:30 08:30 08:50 10.30 09:05
18-Apr-95 02-May-95 23-May-95 06-Jun-95 20-Jun-95 11-Jul-95 18-Jul-95 01-Aug-95 15-Aug-95 29-Aug-95 12-Sep-95 19-Sep-95 26-Sep-95 03-Oct-95 10-Oct-95 17-Oct-95
08:30 09:30 09:00 07:30 07:30 08:10 07:30 08:00 09:00 07:55 09:00 08:00 08:10 10:00 08:30 08:00
10,040 10,125 10,060 10,070 9930 10,040 10,055 10,005 10,060 10,075 20,220 9990 10,060 10,150 9960 10,015
3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0
30.1 30.4 30.0 30.2 29.8 30.1 30.2 30.0 30.0 30.2 60.7 30.0 30.2 30.5 29.9 30.0
5226
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
from 29.8–60.7 m3. The flow rate (3.0 l min1) was calculated from the pressure drop across the critical orifice using a calibration graph provided by the manufacturer.
Deposition velocities for particles >5 mm were calculated assuming negligible gravitational effects using the following formula:
3.3. Field quality assurance/quality control methods
where
To ensure that the meteorological instruments were performing to standard, they were evaluated yearly using manufacturer’s procedures to ensure that they performed accurately and within the manufacturer’s specifications. Field blanks were used to assess sample handling by subjecting them to the same field handling and laboratory analysis as the samples. Sample handling quality assurance protocols were established and documented prior to starting the sampling program to ensure consistent procedures and reduce variability in the data. Personnel were required to wear polyethylene gloves during sample handling.
* Vdð>5 mmÞ ¼ EA uavg ðm s1 Þ;
ð2Þ
EA ¼ the particle effective inertial coefficient ¼ 1:12e30:36=da ;
ð3Þ
where da is the aerodynamic particle diameter * ¼ the mean friction velocity uavg
¼ kuavg ln½Z0 =ðZ dÞ ðm s1 Þ;
ð4Þ
where k=Von Karman’s constant (0.4), Z0 =surface roughness (0.25 m) after Noll et al. (1985) for a mixed institutional, commercial and residential area, Z=height of the wind speed measurement (m), d=height of the sampler above ground (m).
3.4. Laboratory methods Thirty two metals were measured in LPI particulate samples (Table 2). Samples were analysed at ZENON Analytical Services, Vancouver, BC, using inductively coupled plasma atomic emission spectroscopy (ICP) (BCMELP, 1983). The QA/QC component of the analytical analysis consisted of substrate and reagent blanks. Blanks were used to assess reagents, methods, standards, instruments and calibrations.
4. Data processing and analysis 4.1. Concentrations Raw data were blank corrected based on averaged field blank values. Values below the minimum detection limit were set to zero. This was chosen in favor of the 12 detection limit option in order to minimise overestimation error, as the detection limit was high for a number of elements. Concentrations were calculated from blank corrected data by dividing sample weights by the sample volume as reported in Table 1. 4.2. Deposition velocities Deposition velocities (Vd ) for particles p5 mm were calculated after Slinn and Slinn (1980), while those for particles >5 mm were calculated after Noll and Fang (1989). For particles p5 mm, the deposition velocity was calculated using the following formula: Vdðo5 mmÞ ¼ CD uavgðhÞ ðm s1 Þ;
ð1Þ 3
where CD =the drag coefficient (1.3 10 ), uavgðhÞ =the mean wind velocity at height h (m s1).
Table 2 Target analytes for the analysis of trace metals and their detection limits (mg m3) Analyte
Detection limit
Ag Al As B Ba Be Bi Ca Cd Co Cr Cu Fe K Mg Mn Mo Na Ni P Pb S Sb Se Sn Sr Te Ti Tl V Zn Zr
0.00027 0.17 0.0013 0.10 0.0025 0.00050 0.066 1.3 0.15 0.0075 0.050 1.5 1.2 1.0 0.050 0.025 0.010 1.0 0.050 0.10 1.2 0.25 0.35 0.0013 0.075 0.43 0.050 0.17 0.075 0.012 0.43 0.015
5227
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
log transformed and standardised to standard deviation. Metals with >5 non-detectable values were not included in the PCA analysis. Principal components were extracted using a correlation matrix.
4.3. Deposition rates Deposition rates were calculated from the product of the deposition velocity and metal concentration in each particle size fraction. Total deposition (Dt ) was calculated as the sum of deposition rates for each particle size interval. The equation consisted of a combination of the Slinn and Slinn (1980) and Noll and Fang (1989) equations for particles of diameter p5 and >5 mm, respectively X X Dt ¼ ½ð1:3103 Þuavg Cda þ ðda o5 mmÞ
5. Results and discussion 5.1. Meteorological data Meteorological data collected during the study period are presented in Table 3. Results represent average weekly values. Some meteorological data was lost due to data logger overload or adverse weather conditions. The mean temperature ranged from a low of 7.91C to a high of 19.81C. The mean relative humidity ranged from 64.6–92.5% with lower values typical of the summer months. Precipitation varied throughout the study period, ranging from 0.4–89.8 mm with considerably greater amounts of precipitation falling during the fall months. Mean atmospheric pressure values ranged from 1009–1018 mbar. Wind speeds ranged from 4.8– 6.8 km h1. Winds were predominantly from the east during the study period.
ðda >5 mmÞ
h i 2 * ÞCd ð1:12e30:36=da uavg sÞ; a ðng m
ð5Þ
where da =the aerodynamic particle diameter, Cda =the contaminant concentration for particles of aerodynamic diameter ‘‘a’’ (ng m3), uavg¼ mean wind speed (m s1), u avg =mean friction velocity (m s1) 4.4. Principal components analysis Principal component analysis (PCA) was used for source interpretation. Source types were identified by associating the statistically significant component loadings with elements related to specific source types. Data were arranged in a matrix such that each column contained either the concentration of one chemical species or a physical parameter and each row a single weekly sample. Prior to running the analysis, data were
5.2. Metals data Of the 32 metals measured in atmospheric aerosols at Burnaby Lake, 14 were assessed for particle size
Table 3 Meteorological data collected during the sampling perioda Week
Start
Stop
Wind directionb
Wind speed (km h1)c
Temperature (1C)
Pressure (mbar)
RHd (%)
Precipitation (mm)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
11-Apr-95 25-Apr-95 16-May-95 30-May-95 13-Jun-95 04-Jul-95 11-Jul-95 25-Jul-95 08-Aug-95 22-Aug-95 29-Aug-95 12-Sep-95 19-Sep-95 26-Sep-95 03-Oct-95 10-Oct-95
17-Apr-95 28-Apr-95 22-May-95 05-Jun-95 19-Jun-95 10-Jul-95 17-Jul-95 31-Jul-95 13-Aug-95 28-Aug-95 04-Sep-95 18-Sep-95 25-Sep-95 02-Oct-95 09-Oct-95 16-Oct-95
S, SW SW, SE SW SE and S SE, E, W and SW SE SE, NW SW and SE E S E, NE NE NE E and NE E SE
6.2 5.5 5.9 5.9 6.2 5.3 6.7 6.8 6.7 5.1 n/ae n/a n/a n/a 6.1 4.8
7.9 15.1 14.4 16.8 14.9 17.9 19.8 17.4 15.6 15.8 17.0 18.2 15.7 13.2 11.5 11.3
1013 1009 1016 1013 1010 1014 1018 1015 1014 1014 1014 1015 1015 1012 1015 1016
78.1 64.6 69.3 73.9 83.3 78.9 73.1 75.9 85.6 80.1 80.5 82.6 76.0 90.0 90.9 92.5
24.0 0.0 0.4 8.8 9.2 19.6 0.0 51.0 11.2 7.2 9.4 1.6 4.4 33.6 89.8 76.6
a
Values represent means for each sampling period. Prevailing wind direction. c Average wind speed derived from 5 min intervals within the sampling period. d Relative humidity. e n/a denotes data not available due to equipment failure. b
5228
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
distribution, 12 were assessed for total concentrations for the sum of particle size fractions and six were not detected (Table 4). The metals assessed for total Table 4 Summary of assessment of metals in LPI samples from Burnaby Lake Assessed for particle size distributiona
Assessed for total concentrationsb
Not detected
Al As B Ba Ca Fe Mg Mn Na S Sr Ti V Zr
Ag Be Cd Co Cr K Mo P Pb Se Te Zn
Bi Cu Ni Sb Sn Tl
a b
Size fractions quantifiable. Size fractions not quantifiable.
concentrations could not be assessed for individual size fractions due to a high frequency of non-detectable values in the size fractions. Table 5 presents a statistical summary of metal concentrations for the sum of particle fractions. PCA analysis indicated a strong association between aluminum, barium, boron, calcium, iron, magnesium, manganese, sodium, strontium, and vanadium (Fig. 2). Further examination confirmed that most of these metals had a similar time series and particle size distribution. Maximum concentrations occurred during weeks of low precipitation with a large peak for the week including 13 June followed by multiple smaller peaks in the late summer-early fall period (Fig. 3). This is consistent with a westward transport of marine aerosols in the region and elevated concentrations of geological material (Chow and Watson, 1998). Concentration maxima for titanium and vanadium occurred in the late summer and early fall (not shown). While titanium is commonly associated with soil dusts, vanadium is often present in emissions from residual oil combustion (Chow and Watson, 1998). The particle size distribution for this group of metals was mostly dominated by a large peak between 1.7 and 18.4 mm, with a smaller secondary peak at o0.08 mm (Figs. 3 and 4). For the
Table 5 Statistical summary of trace metal concentrations (mg m3) for the sum of particle size fractions (n ¼ 16) Analyte
Detection limit (DL)
Mean
S.D.
Minimum
Maximum
Fraction of data with values 4DL (%)
Ag Al As B Ba Be Ca Cd Co Cr Fe K Mg Mn Mo Na P Pb S Se Sr Te Ti V Zn Zr
0.00027 0.17 0.0013 0.10 0.0025 0.00050 1.3 0.15 0.0075 0.050 1.2 1.0 0.050 0.025 0.010 1.0 0.10 1.2 0.25 0.0013 0.43 0.050 0.17 0.012 0.43 0.015
0.0014 200 0.023 55 0.24 0.0024 480 0.00039 0.0050 0.012 4.8 4.0 48 0.17 0.0018 26 0.45 0.049 6.8 0.0022 7.3 0.010 1.7 0.064 0.084 0.20
0.0038 190 0.068 50 0.30 0.0053 450 0.0012 0.0092 0.032 4.8 2.4 45 0.19 0.0052 21 0.22 0.043 4.3 0.0046 6.3 0.030 0.92 0.069 0.23 0.13
0.002 13 0.0012 3.3 0.0052 0.0004 41 0.0018 0.008 0.025 1.3 1.1 3.6 0.032 0.011 5.0 0.27 0.010 1.6 0.0015 1.2 0.053 0.39 0.032 0.59 0.051
0.015 760 0.28 180 1.1 0.022 1800 0.0045 0.030 0.12 18 8.8 180 0.77 0.018 91 0.89 0.14 18 0.014 26 0.11 3.8 0.23 0.76 0.46
31 100 69 100 100 63 100 13 31 19 81 94 100 88 13 100 100 88 100 25 100 13 100 63 13 100
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
Fig. 2. PCA loadings plot of sample metal concentrations for the sum of particle fractions (n ¼ 16). Factors 1 and 2 explained 53% and 13% of the variability, respectively.
5229
coarse particle fraction the source is likely soil dust and sea salt, while metals in the submicron particle range are likely from various anthropogenic sources. Aluminum, sodium and manganese have been shown to be present in fine particles from diesel and gasoline combustion (Kleeman et al., 2000), calcium and magnesium in lube oil combustion emissions, and submicron particles of iron, magnesium and manganese are emitted from engine wear or fuel components (Graham, unpublished). Manganese is commonly used in the anti-knock agent methylcyclopenpentadienyl manganese tricarbonyl (MMT) in some types of commercial gasolines (Chow and Watson, 1998). Strontium is an alkaline earth metal normally associated with sulphate or carbonate, and boron occurs naturally in soils and as a bi-product of glass manufacturing (Merck Index, 1989). The grouping of the remaining metals in the PCA plot suggested different sources. For example, lead was most closely associated with beryllium (Fig. 2), possibly
Fig. 3. Typical concentration time series for: Al (shown here), B, Ca, Fe, Mg, Mn, Na and Sr for the sum of particle size fractions (solid line, mg m3) and precipitation (dashed line, mm) during sampling period.
Fig. 4. Typical particle size distribution for: Al (shown here), Ba, B, Ca, Fe, Mg, Mn, Na, Sr, Ti and V. Concentrations are expressed as means (mg m37SE), n ¼ 16:
5230
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
reflecting electrical manufacturing sources. Phosphorus and potassium were not associated with any other variables, pointing to unique sources such as agricultural fertilisers for the former and vegetative burning for the latter (Chow and Watson, 1998). Sulphur was associated with wind direction (which was generally from the south, southwest during the two weeks with peak sulphur concentrations (Fig. 5)) indicating marine sources and or inputs from the oil refineries south of the Canada–US border. Sulphur concentration maxima occurred during the spring period (between 11 April and 23 May) and corresponded to a period of low precipitation (Fig. 5). Furthermore, mean annual sulphur concentrations exhibited a bimodal distribution corresponding to coarse and fine particles, respectively (Fig. 6). The fine particle fraction is likely dominated by ammonium sulphate aerosols (Pryor et al., 1997), while the coarse fraction is probably largely composed of sea salt sulphate (Savoy and Prospero, 1982). Seasonally,
sulphur concentrations were highest in the fine particle range during the spring (Fig. 7), suggesting that marine sources, likely in the form of dimethyl sulphide precursors, may play an important role (Kettle et al., 1999). Metal concentrations were generally one to three orders of magnitude higher than those measured 100 km away in rural Abbotsford in 1996 but similar to those measured in urban Taipei, Taiwan (Li et al., 1993). Exceptions were found for aluminum and calcium which were in higher concentrations at Burnaby Lake. Concentrations of the highly toxic metals, arsenic, cadmium and lead, were within current air quality guidelines. Lead was interesting in that it’s time series pattern was similar to that of the crustal elements, suggesting an association with soil dusts. Although lead has been banned from use as a gasoline additive in Canada for over a decade, residual levels are likely to persist in urban particulate matter for many decades to come.
Fig. 5. Sulphur concentration time series for the sum of particle size fractions (solid line, mg m3) and precipitation (dashed line, mm) during the sampling period.
Fig. 6. Particle size distribution of sulphur. Concentrations are expressed as means (mg m37SE), n ¼ 16:
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
Boron exceeded the Ontario Ministry of Environment ambient air quality standard (OME (Ontario Ministry of Environment), 1999) in two of the 16 samples. Deposition velocities ranged between 0.22 and 13 cm s1 and increased with particle diameter (Fig. 8). It should be noted that the errors associated with calculation of deposition velocities are greater for larger width particle fractions, as particle fraction width increases with mean diameter size. This results in larger errors of the estimated deposition rate for the larger particle size fractions. Total deposition rates varied for each metal with mean values ranging between
5231
4.0 mg m2 d1 and 650 mg m2 d1(Table 6). Deposition rates generally reflected concentration patterns with the highest deposition rates occurring for the week including 13 June. Depending on the metal, yearly loadings to the watershed ranged from 90 kg to several thousand tonnes (Table 7). Calcium, aluminum, boron and magnesium had the highest metal loadings to the watershed. Manganese also had relatively high loadings and was reported to exceed federal and provincial water quality guidelines in creeks and rivers within the Burnaby Lake watershed (Sekela et al., 1998), which may support the atmospheric pathway as an important source of this
Fig. 7. Sulphur concentration for the different size fractions during the spring (n ¼ 5), summer (n ¼ 6) and fall (n ¼ 5). Spring: 11 April–20 June ; Summer: 4 July–12 September; Fall: 12 September–17 October.
Fig. 8. Calculated deposition velocities (cm s1) as a function of particle size.
5232
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233
Table 6 Statistical summary of trace metal deposition rates (mg m-2 d1) for the sum of particle size fractions (11 April–10 October 1995) (n ¼ 11) Analytea
Mean
S.D.
Minimum
Maximum
Al As B Ba Be Ca Fe K Mg Mn Na P Pb S Sr Ti V Zr
270,000 100 69,000 260 4.0 650,000 6000 11,000 64,000 350 44,000 920 78 23,000 10,000 2000 61 330
360,000 330 87,000 420 10 870,000 7800 9000 84,000 500 47,000 640 120 12,000 13,000 1400 72 230
10,000 2.1 2900 0.86 0.29 21,000 620 420 2600 6.6 840 270 1.0 8000 850 84 6.6 12
1,300,000 1100 300,000 1100 34 3,100,000 27,000 26,000 300,000 1500 170,000 2300 380 47,000 45,000 4300 180 600
a
Deposition values were not calculated for Ag, Cd, Co, Cr, Mo, Se, Te and Zn due to >50% of the samples being below the detection limit.
Table 7 Estimate of yearly metal deposition rates to entire Burnaby Lake watershed
contribute significantly to the total metal loading to urban watersheds (Chow and Watson, 1998), resulting in an underestimation of the total metal deposition to the Burnaby Lake watershed. Long range transport and deposition likely accounts for a portion of the total atmospheric load to the area, however the magnitude of this contribution is unknown.
6. Conclusion The air pathway appears to be an important source of metals to the Burnaby Lake area. The relatively high metal concentrations measured at the site are a reflection of the highly urbanised location of the lake, receiving inputs from a multitude of natural and anthropogenic sources. The exceedence of the Ontario Ministry of Environment ambient air quality standard for boron in several samples suggests that concentrations may occasionally pose a risk to human health. The relatively high metal deposition rates to the area indicates that the metal contribution from atmospheric sources may represent a significant portion of the total metal load to the Burnaby Lake watershed, while the reported exceedence of water quality guidelines and criteria for several metals in local rivers suggests that aquatic life may be at risk in the more urbanised reaches.
Analyte
Estimated deposition ratea (ton yr1)
References
Al As B Ba Be Ca Fe K Mg Mn Na P Pb S Sr Ti V Zr
5900 2.3 1527 5.7 0.090 14,000 130 240 1400 7.7 980 20 1.7 500 230 45 1.3 7.3
BCMELP (British Columbia Ministry of the Environment, Lands and Parks), 1983. A laboratory manual for the chemical analysis of ambient air emissions, precipitation, soil and vegetation. In: Neil McQuacker (Ed.), Waste Management Branch. Data Standards Group, Vancouver, BC. CEPA/FPAC (Canadian Environmental Protection Act/ Federal Provincial Advisory Committee), 1998. National Ambient Air Quality Objectives for Particulate Matter. Part 1: Science Assessment Document. A report by the CEPA/FPAC Working Group on Air Quality Objectives and Guidelines, ISBN 0-662-63486-1, 19pp. Chow, J. C., Watson, J. G., 1998. Western Washington 1996–97 PM 2.5 Source Apportionment Study. Worldwide Environmental Corporation, 8101 Meadow Vista Dr., Reno, Nevada, 89511. Prepared for Puget Sound Air Pollution Control Agency, 110 Union Street, Suite 500, Seattle, Washington, 98101. Graham, L., Gaseous and Particulate Matter Emissions from In-Use Light Duty Gasoline Motor Vehicles. Environment Canada, Environmental Technology Advancement Directorate, Environmental Technology Centre, Ottawa, Ontario, K1A 0H3, EDM Report #99-67, 90pp (unpublished). Hall, K.J., Anderson, B.C., 1988. The toxicity and chemical composition of stormwater runoff. Canadian Journal of Civil Engineers 15, 98–106.
a
Based on an area of 6060 ha.
metal. It should be noted that due to high background levels in the blanks, loadings shown in Table 7 do not include some of the common elements associated with urban aerosols, such as Si, Cr, Cu and Zn, which can
R. Brewer, W. Belzer / Atmospheric Environment 35 (2001) 5223–5233 Kettle, A.J., Andreae, M.O., Amouroux, D., Andreae, T.W., Bates, T.S., Berresheim, H., Bingemer, H., Boniforti, R., Curran, M.A.J., DiTullio, G.R., Helas, G., Jones, G.B., Keller, M.D., Kiene, R.P., Leck, C., Levasseur, M., Malin, G., Maspero, M., Matrai, P., McTaggard, A.R., Mihalopoulos, N., Nguyen, B.C., Novo, A., Putaud, J.P., Rapsomanikis, S., Roberts, G., Schebeske, G., Sharma, S., Simo, R., Staubes, R., Turner, S., Uher, G., 1999. A global database of sea surface dimethylsulphide (DMS) measurements and a procedure to predict sea surface DMS as a function of latitude, longitude, and month. Global Biogeochemical Cycles 13 (2), 399–444. Kleeman, M.J., Schauer, J.J., Cass, G.R., 2000. Size and composition distribution of fine particulate matter emitted from motor vehicles. Environmental Science and Technology 34 (7), 1132–1142. Li, C.-S., Hsu, L.-Y., Chuang, Y.-Y.T., 1993. Elemental profiles of indoor and outdoor particulate matter less than 10 mm (PM 10) and 2.5 mm (PM 2.5) in Taipei. Chemosphere 27 (11), 2143–2154. Merck Index, 1989. The Merck Index. In: Susan Budavari (Ed.), An Encyclopedia of Drugs and Chemicals, eleventh edition. Merck & Co. Inc. Rahway, NJ, USA, 1606 pp + app. National Research Council, 1993. Protecting Visibility in National Parks and Wilderness Areas. National Academy Press, Washington, DC, 446pp.
5233
Noll, K.E., Pontius, A., Frey, R., Gould, M., 1985. Comparison of atmospheric coarse particles at an urban and nonurban site. Atmospheric Environment 19, 1931–1943. Noll, K.E., Fang, K.Y.P., 1989. Development of a dry deposition model for atmospheric coarse particles. Atmospheric Environment 23 (3), 585–594. OME (Ontario Ministry of Environment), 1999. Summary of Point of Impingement Standards, Point of Impingement Guidelines and Ambient Air Quality Criteria (AAQCs). Standards and Development Branch, Ontario Ministry of Environment, 10pp. Pryor, S.C., Barthelmie, R.J., Hoff, R.M., Sakiyama, S., Simpson, R., Steyn, D., 1997. Reveal: characterizing fine aerosols in the Fraser Valley, BC. Atmosphere-Ocean 35 (2), 209–227. Savoy, D.L., Prospero, J.M., 1982. Particle size distribution of nitrate and sulfate in the marine atmosphere. Geophysical Research Letters 9 (10), 1207–1210. Sekela, M., Brewer, R., Tuominen, T., Sylvestre, S., Moyle G., 1998. Effect of a Rainfall Event on Contaminant Levels in the Brunette River Watershed. Environment Canada, Aquatic and Atmospheric Sciences Division, Environmental Conservation Branch, Pacific and Yukon Region. Vancouver, BC, DOE FRAP 1997-36, 35pp. Slinn, S.A., Slinn, W.G., 1980. Predictions for particle deposition on natural waters. Atmospheric Environment 14, 1013–1016.