Atmospheric Abundances of Nitrogen Species in Rain and Aerosols Over a Semi-arid Region : Sources and Deposition Fluxes

The temporal variation for a three-year period (2000-2002) in the atmospheric abundances of principal nitrogen species (NH4 and NO3) has been studied in rain and aerosols from an urban city (Ahmedabad, 23.0N, 72.6E) located in a semi-arid region of western India. Their concentrations in ambient aerosols over the annual seasonal cycle exhibit large variation [NH4: < 0.001 to 1.3 (GM = 0.25 μg/m); NO3: 0.09 to 4.4 (GM = 1.3 μg/m)]; with systematically higher concentrations during Nov-Feb (drier period) and relatively low during Jun-Aug (wet season). In comparison, abundances of NH4 and NO3 in individual precipitation events (n = 91) collected during the southwest monsoon (Jun-Aug) for three years varied as NH4: <1.0 to 220 (VWM: 30 μeq/L) and NO3: 1.3 to 115 (VWM: 13 μeq/L). Using corresponding rainfall data, the wet-deposition fluxes of NH4 during 2000, 2001 and 2002 have been derived: 364, 327 and 297 mg/m/y, respectively; which are considerably higher than the dry-deposition fluxes (16, 11 and 16 mg/m/y). The wet deposition of NO3 over the three years varied as 392, 569 and 487 mg/m/y in comparison to dry-deposition fluxes averaging as 201, 134 and 137 mg /m/y. These results also indicate that both dry and wet deposition of NO3 accounts for as much as ~65% of the total inorganic-N species (NH4 + NO3), possibly mediated via in-situ chemical reactions of acidic species (NO3 and its precursors) with mineral aerosols. Such regional scale studies are important for quantitative assessment of atmospheric deposition of biogeochemically important trace species and bring to focus the role of semi-arid regions where current knowledge is limited.


INTRODUCTION
Nitrogen is a major nutrient in terrestrial ecosystem and an important catalyst in tropospheric photochemistry.Atmospheric nitrogen compounds affect Earth's radiation budget, acidify ecosystems, and cause degradation and eutrophication of lakes, estuaries and coastal oceanic regions (Galloway, 1995;Galloway et al., 2004;Holland et al., 2005).Among inorganic nitrogen species, NH 4 + , NO 3 -and their precursors are the most dominant species present in the atmosphere.
In the present-day scenario, the major sources of atmospheric NH 4 + , NO 3 -and their precursors (NH 3 and NO x ) are attributed to vehicular emissions, biomass and fossil-fuel burning, human and animal excreta, microbial decomposition of biomass etc.The typical residence time of nitrogen species in the atmosphere of about a week is similar to that of ambient aerosols.During their long-range transport, these species undergo a variety of physico-chemical transformation processes and are ultimately removed from the atmosphere via dry and wet deposition.The wet deposition is considered to be the major removal pathway of N-species, accounting for > 80% in many regions.However, in arid and semi-arid regions, the dry deposition can be substantial; but the relative significance of dry and wet deposition processes are poorly documented.In this context, regional-scale programs are the key components of global nitrogen budget and to predict the effects of nitrogen cycle on climate change (Galloway et al., 2004;Holland et al., 2005).The present study reports on the atmospheric abundances of NH 4 + and NO 3 -in rain and aerosols, their seasonal variability and deposition fluxes (dry and wet) from an urban area located in a semi-arid region of western India.

Site description
Our sampling site (located in Ahmedabad city: 23.0 o N, 72.6 o E, 49 m asl, population ~5.5 million) represents a typical urban atmosphere of a semi-arid region in western India.To its northwest lies the great Thar Desert.In the southwest is the Arabian Sea and to its northeast is the continental area inhabited by a large human population and diverse industrial activities (Fig. 1).
The contrasting wind patterns during Nov to Feb (representative of NE winds from large-scale pollution sources that affect the air quality) and May to Aug (southwesterlies transporting sea-salts) dictate the aerosol composition over the study site (Rastogi, 2005).The wind fields during Mar-Apr and Sept-Oct represent transition phase in the circulation pattern.The annual rainfall averages ~650-700 mm, all of which occurs during the SW monsoon (Jun-Aug).The

Sampling and analyses
The sampling gear for aerosol and rainwater collection was set-up at a height of ~40 m above the ground level in order to avoid near-surface local sources.Bulk-aerosols (n = 99) were collected, from Jun 2000 to Jan 2003, onto clean precombusted tissuquartz filters (PALLFLEX TM , 20 × 25 cm) by operating a high-volume air sampler (flow rate: ~1.2 m 3 /min).On an event basis, "wet-only" precipitation samples (n = 91) were also collected during the SW monsoon (Jun-Aug) for three successive years (2000)(2001)(2002).The total rainfall during 2000, 2001 and 2002 was 730, 680 and 400 mm, respectively; all of which occurred during the SW monsoon.The rainwater samples and water extracts of aerosols were analyzed for NH 4 + and NO 3 -along with other major ions (Na + , K + , Mg 2+ , Ca 2+ , Cl -, SO 4 2-and HCO 3 -) by standard techniques.Further details on analytical procedures are found in Rastogi and Sarin (2005) and Rastogi (2005).The cation/anion ratio (on equivalent basis) in water-extracts of the aerosols ranged from 0.80 to 1.14 (average: 1.02, sd: 0.08).The chemical data in precipitation events also follows the 1:1 trend, except in a few heavier precipitation events when the solute concentration (cation + anion) is less than 100 μeq/L.

Source apportionment for NH 4 + and NO 3 -in aerosols and rainwater
In order to decipher the dominant sources of NH 4 + and NO 3 -in aerosols and precipitation events, a linear regression analysis was performed among various ions measured in water-extracts of aerosols and in rainwater samples (Fig. 2).We have considered nss-SO 4 2-as a constituent of anthropogenic origin, abundance of nss-Ca 2+ signifies its dominant contribution from mineral dust, and that of nss-K + from biomass burning sources (Rastogi, 2005).Unlike rain events occurring during SW monsoon (Jun-Aug), the chemical data on aerosol samples (collected around the year) has been split into three seasons based on prevailing wind patterns (Fig. 2).
Among the water-soluble aerosol species, a significant linear co-variance (r 2 = 0.38) is observed between nss-SO 4 2-and NO 3 -(excluding the wet season data, Jun-Aug), suggesting a contribution of NO 3 -from anthropogenic sources.Furthermore, linear regression parameters for nss-K + versus NO 3 -scatter plot (slope = 1.54 and r 2 = 0.45, Fig. 2) addresses the issue that biomass burning could also be a dominant source of NO 3 -as KNO 3 (expected NO 3 -/K + wt ratio = 1.59, based on stoichiometry).However, a significant intercept (c = 0.98) in the above plot would imply NO 3 -contribution from anthropogenic sources as discernible from co-variance between nss-SO 4 2-and NO 3 -(Fig.2).Thus, multiple sources can be invoked for atmospheric NO 3 -aerosol.
The scatter plot between nss-Ca 2+ and NO 3 -also exhibit a linear trend (except for a few samples marked by high abundance of Ca 2+ , > 4 μg/m 3 ) (Fig. 2).This is attributed to neutralization reaction between HNO 3 and mineral dust (CaCO 3 ) (Laskin et al., 2005;Rastogi, 2005).The outlier (Ca 2+ > 4 μg/m 3 ) corresponds to aerosol samples collected during high-dust events.It is noteworthy that regression analysis of NH 4 + with SO 4 2-shows a large scatter (Fig. 2); the lowest concentrations of NH 4 + occurring during the May to Aug (a period dominated by SW winds of marine origin), and highest during Nov-Feb (NE winds transporting pollutants from large continental areas) (Fig. 3).A linear trend observable between NH 4 + and nss-SO 4 2-for the data set with NH 4 + abundance exceeding 0.5 μg/m 3 (Nov-Feb, Fig. 2) suggests its occurrence as sulphate salts.A large seasonal variation in the abundance of aerosol NH 4 + is attributed to its partitioning between gaseous and particulate phase; the contribution from the latter phase is relatively more pronounced during winter months (Nov-Feb) under the influence of NE winds.In the case of precipitation samples, the regression analysis for NH 4 + and NO 3 -with K + , Ca 2+ and SO 4 2-exhibits characteristics trends somewhat similar to those observed for water-soluble species of aerosols (Fig. 2).

Seasonal and inter-annual variation of NH 4 + and NO 3 -in aerosols and rain
The abundances of NH 4 + and NO 3 -measured in aerosols and precipitation samples exhibit large variations (Fig. 2).The aerosol abundance of NH 4 + ranges from < 0.001 to 1. + concentration often below detection limit (0.001 μg/m 3 ) (Fig. 3).During May-Aug, relatively low abundances of nitrogen species are attributed to the predominance of SW winds (marine air mass), as well as rapid removal of aerosols (via wet deposition).The volatilization loss of NH 4 + (if any) present as NH 4 NO 3 or NH 4 Cl aerosols could also be significant during Apr-Jun before the onset of SW monsoon when ambient temperature over the study site reaches maximum [a mechanism similar to that suggested by Matsumoto and Okita (1998)].Relatively high concentrations of both species are observed during Nov-Feb (Fig. 3), under the influence of NE winds that dominate the transport of continental pollutants (Rastogi, 2005).
The volume-weighted-mean (VWM) concentrations of NH 4 + and NO 3 -in precipitation events collected during SW monsoon, for three years, are shown in Fig. 4. The abundance of NH 4 + ranges from < 1.0 to 220 μeq/L (VWM: 30 μeq/L) and accounts for nearly 20% to the cation balance (Rastogi and Sarin, 2005).The relatively low NH 4 + concentration (close to detection limit) in aerosols during May-Aug (Fig. 3) and its significant abundance in precipitation events during SW monsoon period ascertains its dominant occurrence in gaseous phase over the study site.The abundance of NO 3 -ranges from 1.3 to 115 (VWM: 13 μeq/L) and on average contributes ~10% to the anion balance (Rastogi and Sarin, 2005).The large variability observable for NH 4 + and NO 3 -in rain events is attributed to the variations in the source strength, antecedent dry period between the successive rain events and precipitation amount.A successive increase in VWM of NO 3 -over three-years (Fig. 4) can be viewed as systematic increase in NO x emission over the Asian region (Van Aardenne et al., 1999;Streets et al., 2003).The VWM of NH 4 + does not show a significant and systematic inter-annual variability.Where, [X]: Average concentration of a species X (mg/m 3 ); V d : deposition velocity (m/y) for that species; C: VWM concentration of species (mg/m 3 ); P: Precipitation (m/y).V d is considered as 0.1 cm/s for NH 4 + and 0.3 cm/s for NO 3 - (Duce et al., 1991).
A comparison of NH 4 + and NO 3 -abundances over different sites in India, as well as from a few global regions, are presented in Table 1, thus representing their dry deposition fluxes (assuming uniform deposition velocities for NH 4 + and NO 3 -over all the sites).Their abundances over Ahmedabad are similar to those reported over Christchurch (New Zealand) and Dundee (United Kingdom) while significantly lower than those reported over other sites in India and the world (Table 1).The observed differences are attributable to their different source inputs, as well as source strength and meteorological conditions.The wet deposition fluxes of NH 4 + and NO 3 -over Ahmedabad are comparable to those reported over Pune (Pillai et al., 2001), Delhi (Jain et al., 2000), Hyderabad (Kulshreshtha et al., 2003), Amersfoort and Louis Trichardt (Mphepya et al., 2004), Banizoumbou (Galy-Lacaux and Modi, 1998) and Ankara (Kaya and Tuncel, 1997); but significantly lower than those at Reston (Conko et al., 2004), Sau Paulo (Rocha et al., 2003) (Fig. 6).et al., 2002; b Momin et al., 1999; c Parashar et al., 2001  d Kulshrestha et al., 1998; e Salam et al., 2003; f Kocak et al., 2004;   g Yao et al., 2002; h Wang and Shooter, 2001; i Chow et al., 1993;   j Qin and Oduyemi, 2003.
monthly average temperature from Jan to Dec varies between 21 and 34 o C with relatively high temperatures (32 to 34 o C) during summer months (Apr-Jun) and low (21 to 23 o C) during winter months (Dec-Feb).

Fig. 2 .
Fig. 2. Source apportionment of NH 4 + and NO 3 -in rain and aerosols based on scatter plots with major ions.Encircled points are not included in regression analyses.

Fig. 3 .
Fig. 3. Seasonal average abundances of NH 4 + and NO 3 -in aerosols exhibits relatively high values during Nov-Feb due to dominance of polluted continental air-mass (NE winds) and low during May-Aug (SW winds).Error bar represents the standard deviation.

Fig. 4 .-
Fig. 4. Volume-weighted-mean (VWM) concentration of NH 4 + and NO 3 -in rain events during SW monsoon shows linear increase in NO 3 -abundance, attesting to predicted trends over Asia.

Fig. 5 .
Fig. 5. Dry and wet deposition fluxes of NH 4 + and NO 3 -over an urban area (Ahmedabad).Wet deposition seems to be the major removal pathway for both the species and inter-annual variability in their deposition fluxes is insignificant.

Fig. 6 .
Fig. 6.Wet deposition fluxes of NH 4 + and NO 3 -over Ahmedabad (3-year average) and comparison with a few other study sites.By and large, wet fluxes are comparable to those reported over other sites.

Table 1 .
Abundances (μg/m 3 ) of NH 4 + and NO 3 -in aerosols over Ahmedabad and a few sites in India and the world.
a Venkatraman