Al High
Resume:
Characterization of nutrients in the atmospheric wet
and dry deposition observed at the two monitoring sites
over Yellow Sea and East China Sea
Guosen Zhang & Jing Zhang & Sumei Liu
Received: 28 July 2005 / Accepted: 9 February 2007 /
Published online: 14 March 2007
# Springer Science + Business Media B.V. 2007
Abstract To investigate the atmospheric deposition of nutrients into the coastal and shelf
regions of the northwest Pacific Ocean, observation sites were established upon Qianliyan
Island (within the Yellow Sea) and the Shengsi Archipelago (within the East China Sea),
respectively. Nutrient concentrations, including NH ; NO ; NO ; PO3 and SiO2, were
3 2 4 3
4
determined in both aerosols and rainwater samples. The analytical results contain clear
seasonal signatures, with high values during the dry season and low values during the rainy
season. Similar trends are observed for deposition fluxes. The amount of wet deposition is
greater than that of dry deposition for the studied nutrient species. The influence of
meteorological factors such as rainfall means that samples from Qianliyan Island record
higher nutrient values than those from Shengsi. Along with riverine inputs, atmospheric
deposition plays an important role in determining the biogeochemistry of nutrient species in
coastal and shelf oceans.
Keywords Dry deposition . East China Sea . Nutrient . Yellow Sea . Wet deposition
1 Introduction
It has been increasingly recognized over the past 20 years that atmospheric deposition is an
important pathway by which nutrients such as nitrogen and phosphorus are delivered to the
surface oceans (Duce et al. 1991; Enell and Fejes 1993; Gao and Duce 1997; Herut et al.
1999a, b; Jickells 1995; Prospero and Savoie 1989; Spokes et al. 2000; Whitall et al.
2003). The effect of atmospheric inputs of nutrients on biological cycles is particularly
G. Zhang : J. Zhang : S. Liu
College of Chemistry and Chemical Engineering, Ocean University of China,
5 Yushan Road, Qingdao, 266003, People s Republic of China
G. Zhang : J. Zhang
State Key Laboratory of Estuarine and Coastal Research, East China Normal University,
3663 Zhongshan North Road, Shanghai, 200062, People s Republic of China
e-mail: *******@*****.****.***.**
J Atmos Chem (2007) 57:41 57
42
important for oligotrophic oceanic provinces (Migon and Sandroni 1999; Owen et al.
1992), and episodic deposition may even induce algal blooms (Zhang 1994; Zhang and
Liu 1994).
The Yellow Sea is a semi-closed area of the Northwest Pacific Ocean with a surface area
of 380 109 m2 and average depth of 44 m. The sea is surrounded by areas of rapid
population growth and economic development in both China and Korea. As no large rivers
discharge directly into the Yellow Sea, atmospheric deposition is an important source of
nutrients in this region, especially for the central part of the sea (Zhang et al. 1992). The
East China Sea lies over the broad shelf of the Northwest Pacific Ocean, with a surface area
of 770 109 m2 and average depth of 370 m. The Changjiang (Yangtze River) carries a
large amount of fresh water (928.2 109 m3/y) to the Pacific Ocean, along with abundant
nutrient compounds (Liu et al. 2003).
The Yellow Sea and East China Sea are adjacent to the East Asia mainland and are
influenced by a monsoonal climate and strong emissions of natural and anthropogenic
compounds into the atmosphere. Soil-derived dust and anthropogenic compounds are
transported over the Yellow Sea and East China Sea via atmospheric circulation (Chen et al.
1997; In and Park 2002; Zhang et al. 2001), while aerosols from the ocean also play an
important role in the chemical characteristics of precipitation (Lee et al. 2000). Limited
data are available on nutrients in atmospheric deposition recorded from island observations
in the northwest Pacific Ocean (Zhang and Liu 1994; Zhang et al. 1999).
The present study focuses on the composition of nutrients in atmospheric wet and dry
deposition and their seasonal variability. We took measurements of such deposits at island
observation sites in the Yellow Sea and the East China Sea to understand: (1) the factors
that regulate aerosol concentrations and rainwater chemistry, and (2) the role of atmospheric
nutrient deposition, compared with other sources, in the biogeochemistry of shelf regions of
the Northwest Pacific Ocean (i.e., the East China Sea and the Yellow Sea).
2 Materials and methods
2.1 Description of sampling sites
Rainwater and aerosol samples were collected from Qianliyan Island within the Yellow Sea
and the Shengsi Archipelago within the East China Sea (Fig. 1). As described in Table 1,
Qianliyan Island is located in the northwest of the Yellow Sea, and has an elevation of
75 m. In winter and spring, the air mass over the Yellow Sea generally comes from the
northwest, i.e., the mainland, whereas in summer the influence of the southeast monsoon is
dominant (Zhang et al. 2003). The Shengsi Archipelago is located in the western East
China Sea, approximately 70 km southeast of Shanghai. For sampling in the East China
Sea, rainwater was collected from Huanglong Island within the Shengsi Archipelago, at an
elevation of 224 m. Over the Shengsi Archipelago (herein Shengsi), north to northeast
winds are dominant during winter and early spring, whereas east and southeast winds are
dominant during summer and fall (Zhang et al. 2003). At both sites, rainwater collectors
were co-located with meteorological observatories.
2.2 Rainwater sampling
Rainwater samples were collected using a Plexiglas gauge of 400-cm2 horizontal area. The
samplers were carefully cleaned with dilute HCl and rinsed with Milli-Q water before use.
J Atmos Chem (2007) 57:41 57 43
Fig. 1 Locations of island-based
sampling sites used in the present
study: Qianliyan Island within the
Yellow Sea and Shengsi within
the East China Sea. Refer to
Table 1 for detailed descriptions
of these sites
The samplers were opened immediately prior to precipitation and closed again once the rain
had stopped. This procedure minimizes the influence of dry deposition. Meteorological
information was simultaneously recorded. Occasionally, when two consecutive rain events
occurred within a single day, the gauges were left open between the two events. In the
laboratory, rainwater samples were filtered on a clean bench using 0.45- m filters
precleaned with acid (HCl); the filtrates were poisoned with CHCl3 and preserved at low
temperatures until chemical analysis. Sampling blanks were estimated by adding 200 ml of
Milli-Q water to a precleaned rain collector for 1 2 days and then analyzing for nutrients
Table 1 Description of sampling sites
Site Site description
Located on an island situated 50 60 km from the mainland, with 20 30 habitants.
Qianliyan Island
(Yellow Sea) The rain collector is housed in a meteorological observatory. Local air pollution is
derived from the nearby fishing harbor and passing boats, as well as anthropogenic
sources from the mainland to the north and west, which has high densities of
population and manufacturing.
Shengsi Located on an island situated 60 km from the mainland, with 11,000 habitants over
an area of 5.12 km2. Local air pollution is derived from a nearby fishing harbor,
(East China Sea)
anthropogenic emissions, and remote pollution from the mainland to the northwest
and southwest.
J Atmos Chem (2007) 57:41 57
44
(Zhang et al. 1999). The results indicated that the sampling blanks were low, accounting
for less than 3% of the sample values; this value was easily removed by subtraction from
measured values.
2.3 Aerosol sampling
Low-volume pumps were used to collect aerosol samples on 0.40- m pore-size Nuclepore
filters at a pumping rate was 120 l/min of air. The inlet diameter of the filter holder was
45 mm to allow laminar airflow and minimize collisions of aerosol particles in the sampling
chamber. We didn t use any wind direction control on the pump due to logistical difficulties
with establishing such a system during fieldwork. The filters and sampler holders were
cleaned and dried in a calss-100 lamina-airflow bench. The sample holders were then
attached to the end of clean bamboo poles ( 4 5 m), and pumps were installed 10 m
downwind and/or isolated from the sampling system to avoid contamination. A sampling
duration of 24 h was selected according to weather conditions and pumping rate. The flow
meter was calibrated before use and checked again after the sampling campaign. We found
that an overloaded filter reduces the flow rate, especially when humidity is high (e.g., 80
90%). In such a case, sampling was suspended until a new filter was fitted. The blanks for
sampling were obtained by periodically exposing filters in clean samplers for 24 h without
pumping. Filters were changed in clean plastic tents with adequate protection and precautionary
measures (e.g., lab coat and plastic gloves) necessary to avoid contamination. Laboratory
manipulations were carried out in a class-100 lamina-airflow bench. All laboratory wares used
for sample collection and treatment were pre-cleaned with diluted HCl, rinsed with Milli-Q
water, and dried before use (Liu et al. 2002; Zhang et al. 2002). The aerosol sampling
efficiency for particles with a diameter of 35 m is estimated to be 50% (Friendlander
2000; Liu et al. 2002).
2.4 Analytical methods
In the laboratory, the Nuclepore filters were dried at 40 45 C in an oven for 24 h, followed by
conditioning in a dessicator for 8 h at room temperature within a balance room. The filters
were then reweighed to determine the amount of total aerosols. To determine the water-
soluble fraction of inorganic nutrients (i.e., NH ; NO ; NO ; PO3 ; and SiO2 in the
3 2 4 3
4
aerosols, 5 ml Milli-Q water was added and the samples were ultrasonically extracted for
30 min (Model: KS-3000). The extraction solutions were filtered using precleaned 0.45- m
pore-size filters. Blanks for the extracting method were estimated from unused filters
treated in the same way.
Nutrients in rainwater and aerosols extractions were determined using an autoanalyzer
(Model: Skalar SAN PLUS) (Grasshoff et al. 1983; Zhang et al. 2003). The precision of the
chemical analysis is provided in Table 2. The quality of ionic analyses was checked daily
by repeated determinations for selected samples; this yielded a standard deviation of
relative to the average. The analysis of nutrient species involved an inter-laboratory
comparison with the University of Hamburg, Germany, which showed satisfactory results
Table 2 Precision of chemical analyses (n =7)
SiO2
NO NO PO3
NH
3 2 4 4 3
Precision 1.34 1.81 2.71 2.04 2.81
J Atmos Chem (2007) 57:41 57 45
with differences of 10 20%. To assess the source of aerosols, air mass trajectory analysis
(NOAA Air Resources Laboratory, HYSPLIT model, FNL dataset) was also undertaken.
3 Results
3.1 Dry deposition
We analyzed a total of 174 aerosol samples collected from Shengsi during the period April
1999 July 2003 and 155 samples collected from Qianliyan Island during the period
February 1999 October 2003. As shown in Table 3, concentrations of aerosol particles
were determined from the sampling volume and the weight difference in Nuclepore filters.
Seasonal variations were assessed from sample groups for spring (March May), summer
(June August), autumn (September November), and winter (December February).
For samples from Qianliyan Island, aerosol concentrations varied over a wide range
from 0.161 g/m3 in autumn to 256 g/m3 in spring: a seasonal difference of between two
and three orders of magnitude. In comparison, aerosol concentrations recorded at Shengsi
varied within a relative limited range from 3.01 g/m3 in autumn and summer to 167 g/m3
in spring (Table 3). At Qianliyan Island, the seasonal average concentration of aerosols in
spring is twice as high as that in other seasons; a similar distribution was also observed at
Shengsi. The mean concentration of aerosols at Qianliyan Island decreased in the order of
spring > winter summer > autumn, while at Shengsi the concentration changed in the
decreasing order of winter spring > autumn > summer. Higher concentrations were
recorded during spring than in summer and autumn because of the frequent occurrence of
dust storms in the East Asian inland desert and Gobi Desert and the passage of northwest
and/or northerly air masses originating from Siberia. With the movement of these air
masses, large volumes of soil dust was entrained and transported to coastal regions of the
Yellow Sea and East China Sea, thus generating increased concentrations of aerosols
(Zhang et al. 2002). During summer and autumn, air masses mainly originated from the
tropical Pacific Ocean under the influence of the summer monsoon. During this time,
marine aerosols are a major source of aerosols. In addition, during the sampled years,
rainfall amount and rainfall frequency were both high during summer and fall; hence, a
washing-out effect was very important during these years. The residence time of aerosols is
expected to be longer in winter than in summer when there are fewer opportunities for
material aerosols to be accumulated in the atmosphere.
3.1.1 Seasonal variations
As aerosols in the atmosphere originate from multiple sources, the composition of nutrients
changes considerably over the four seasons. Nitrate, NO ; and NH are defined as
2 4
Table 3 Concentrations of atmospheric aerosols at Qianliyan Island and Shengsi ( g/m3)
Site Spring Summer Autumn Winter Annual average
10.9 256 5.19 124 0.161 150 9.57 37.3
Qianliyan Island Range
Average 53.5 25.1 22.3 25.7 31.3
4.97 167 3.01 78.8 3.01 82.5 3.70 128
Shengsi Range
Average 20.8 12.2 16.9 21.9 17.6
J Atmos Chem (2007) 57:41 57
46
secondary aerosol-associated species and are not associated with primary aerosols (Chen
et al. 1997, 2004). The combustion of fossil fuels is a significant source of NOx, whereas
NH may originate from anthropogenic emissions such as animal waste and the
4
application of chemical fertilizers (Whitall et al. 2003). PO3 and SiO2 originate from
4 3
anthropogenic, continental, and rock/soil (i.e., natural) sources (Herut et al. 1999a).
Figure 2 shows seasonal variations in nutrient concentrations within the aerosols (mol/kg,
dry weight). At both sites, aerosols recorded high values of NO NO in winter due to the
3 2
combustion of fossil fuels and predominant northwest wind from the mainland, whereas low
values were recorded in summer because of washing-out associated with frequent rain events
(Fig. 2). In spring, soil dust is the major source of aerosols because of frequent dust storms at
this time of year over the East Asian mainland and the movement of air masses toward the
mid-latitude Pacific Ocean; consequently, high concentrations of aerosols are recorded in
spring (see Table 3). We also observed relatively low values of NO NO ; NH 3 2 4
and PO3 mol=kg in aerosols during spring. Similar trends have been observed for
4
atmospheric samples collected from the Southeast Mediterranean, where the concentration
of nutrients in aerosols dropped to 1/160 1/64 of normal levels during dust storm events
(Herut et al. 2002). Most of the silicate and phosphate in dust is not readily dissolved in
Milli-Q water (Herut et al. 2002). In summer, high temperatures and the related
decomposition of biomass leads to large emissions of NH3; consequently, NH in aerosols
4
Fig. 2 Seasonal variations in nutrients within aerosols (mol/kg) based on seasonal divisions of spring
(March May), summer (June August), autumn (September November), and winter (December February) at
Qianliyan Island and Shengsi. The error bars represent analytical errors
J Atmos Chem (2007) 57:41 57 47
also shows high values for this period. PO3 and SiO2 concentrations peak during summer;
4 3
the reason for this needs to be investigated in future work.
A comparison of the nutrient composition of aerosols at the two sample sites of the
present study indicates that aerosols at Qianliyan Island have higher levels of
SiO2 and PO3 but lower levels of NO NO and NH . The Qianliyan Island site might
3 4 3 2 4
therefore receive greater levels of dusts, which are known to be a major source of
phosphorus and silicate, whereas Shengsi is more affected by local and remote
anthropogenic emissions such as NOx and NH3.
3.2 Ionic composition of rainwater samples
A total of 198 rainwater samples were collected at Shengsi from May 2000 to April 2003,
and 75 samples were recovered at Qianliyan Island from April 2000 to March 2003. The
analytical results are listed in Table 4. The rain volume-weighted average concentration (C )
of chemical species can be calculated using the following equation:
P
n
Ci Qi
i 1
C 1
P
n
Qi
i 1
where Ci is the ion concentration and Qi is the rainfall for the ith precipitation event.
Concentrations of the studied species in atmospheric wet deposition are highly variable.
Absolute levels of concentrations vary over one to five orders of magnitude at both sample
sites, depending on the species of interest. In general, most of the high nutrient-
concentration events occurred during low rainfall ( 1 m diameter) of
aerosols, whereas ammonium is predominant in the fine fraction (
Contact this candidate