1   Submitted to ACPD: Paper acp-2011-970 2   3   Nitrogen deposition to the United States: distribution, sources, and processes 4   5   Lin Zhang1,2, Daniel J. Jacob1,2, Eladio M. Knipping3, Naresh Kumar4, J. William 6   Munger1,2, Claire C. Carouge2, Aaron van Donkelaar5, Yuxuan Wang6, Dan Chen7 7   8   [1] {Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA, 9   USA} 10   [2] {School of Engineering and Applied Sciences, Harvard University, Cambridge, MA, 11   USA} 12   [3] {Electric Power Research Institute, Washington, DC, USA} 13   [4] {Electric Power Research Institute, Palo Alto, CA, USA} 14   [5] {Department of Physics and Atmospheric Science, Dalhousie University, Halifax, 15   Canada} 16   [6] {Ministry of Education Key Laboratory for Earth System Modeling, Center for Earth 17   System Science, Institute for Global Change Studies, Tsinghua University, Beijing, 18   China} 19   [7] {Department of Atmospheric and Oceanic Sciences, University of California, Los 20   Angeles, CA, USA} 21   22   Correspondence to: Lin Zhang ([email protected]) 23   24   25   26   27   28   29   30   31     1 32   Abstract 33   34   We simulate nitrogen deposition over the US in 2006-2008 by using the GEOS-Chem 35   global chemical transport model with 1/2° × 2/3° horizontal resolution over North 36   America and adjacent oceans. US emissions of NO and NH in the model are 6.7 and 2.9 x 3 37   Tg N a-1 respectively, including a 20% natural contribution for each. Ammonia emissions 38   are a factor of 3 lower in winter than summer, providing a good match to US network 39   observations of NH (≡ NH gas + ammonium aerosol) and ammonium wet deposition x 3 40   fluxes. Model comparisons to observed deposition fluxes and surface air concentrations 41   of oxidized nitrogen species (NO ) show overall good agreement but excessive y 42   wintertime HNO production over the US Midwest and Northeast. This suggests a model 3 43   overestimate N O hydrolysis in aerosols, and a possible factor is inhibition by aerosol 2 5 44   nitrate. Model results indicate a total nitrogen deposition flux of 6.5 Tg N a-1 over the 45   contiguous US, including 4.2 as NO and 2.3 as NH . Domestic anthropogenic, foreign y x 46   anthropogenic, and natural sources contribute respectively 78%, 6%, and 16% of total 47   nitrogen deposition over the contiguous US in the model. The domestic anthropogenic 48   contribution generally exceeds 70% in the east and in populated areas of the west, and is 49   typically 50-70% in remote areas of the west. Total nitrogen deposition in the model 50   exceeds 10 kg N ha-1 a-1 over 35% of the contiguous US. 51   52   53   54   55   56   57   58   59   60   61   62     2 63   1. Introduction 64   65   Atmospheric inputs of reactive nitrogen (fixed nitrogen) to ecosystems have increased by 66   more than a factor of 3 globally due to human activity, significantly perturbing the global 67   nitrogen cycle (Vitousek et al., 1997; Galloway et al., 2004). Adverse effects may include 68   soil acidification (Bowman et al., 2008), eutrophication (Bouwman et al. 2002), and a 69   reduction in plant biodiversity (Stevens et al., 2004). Increased nitrogen deposition may 70   enhance CO uptake by the land and ocean, though the climate benefit would be offset by 2 71   associated N O emission (Reay et al., 2008). The US Environmental Protection Agency 2 72   (EPA) is presently developing secondary air quality standards for protection of 73   ecosystems against the detrimental effects of nitrogen deposition (US EPA, 2008). This 74   requires a better understanding of nitrogen deposition over the US in its various forms 75   and including contributions from sources both natural and anthropogenic, foreign and 76   domestic. We use here a nested version of the global GEOS-Chem chemical transport 77   model (CTM) to address these issues. 78   79   The anthropogenic contribution to nitrogen deposition is mainly driven by emissions of 80   fixed nitrogen including nitrogen oxide radicals (NO ≡ NO + NO ) and ammonia (NH ). x 2 3 81   These species also have natural sources. NO is emitted to the atmosphere by x 82   combustion, microbial activity in soils, and lightning. In the atmosphere, NO is oxidized x 83   to nitric acid (HNO ) and organic nitrates on a time scale of less than a day. These 3 84   different forms can be deposited to ecosystems by direct uptake (dry deposition). In 85   addition, HNO is highly soluble in water and is scavenged efficiently by precipitation 3 86   (wet deposition). NH is a major component of nitrogen cycling through the biosphere. It 3 87   is emitted to the atmosphere by agriculture (mostly animal husbandry and fertilizer use), 88   natural terrestrial and marine ecosystems, and fires. NH in the atmosphere can combine 3 89   with H SO (from SO oxidation) and HNO to produce ammonium sulfate and nitrate 2 4 2 3 90   particles. Dry deposition is fast for gaseous NH but slow for ammonium particles, while 3 91   wet deposition is efficient for both. 92     3 93   The lifetime of fixed nitrogen in the atmosphere is sufficiently short that most of the 94   nitrogen deposition for a large country such as the US is expected to be of domestic 95   origin. However, transboundary transport including on intercontinental scales can also be 96   significant (Asman et al., 1998; Dentener et al., 2006; Sanderson et al. 2008). A number 97   of studies have estimated an export efficiency of 20-30% for nitrogen oxides (NO ≡ NO y x 98   and its oxidation products) emitted in the US (Jacob et al., 1993; Kasibhatla et al., 1993; 99   Liang et al., 1998; Li et al., 2004). No analysis has been conducted so far on the relative 100   contributions from domestic, foreign, and natural sources to the different forms of 101   nitrogen deposition over the US. 102   103   Here we use a nested continental scale version of the GEOS-Chem global CTM (Y. 104   Wang et al., 2004; Chen et al., 2009) with horizontal resolution of 1/2° × 2/3° over North 105   America and 2° × 2.5° for the rest of the world. Three-year GEOS-Chem simulations for 106   2006-2008 are conducted to quantify the sources and processes for nitrogen deposition to 107   the US. We present an extensive evaluation for 2006 with surface measurements of wet 108   deposition fluxes, HNO and aerosol concentrations, and satellite observations of NO . 3 2 109   We quantify the contributions to nitrogen deposition from wet vs. dry processes and from 110   individual nitrogen species. We also separate the contributions from domestic 111   anthropogenic, foreign anthropogenic, and natural sources. 112   113   2. The GEOS-Chem nested-grid model 114   2.1 General description 115   116   The GEOS-Chem 3-D global model of atmospheric composition (v8-02-03; http://geos- 117   chem.org) was originally described by Bey et al. (2001) and Park et al. (2004). The 118   model here is driven by GEOS-5 assimilated meteorological data for 2006-2008 from the 119   NASA Global Modeling and Assimilation Office (GMAO). The data are available with a 120   temporal resolution of 6 hours (3 hours for surface variables and mixing depths), a 121   horizontal resolution of 1/2° latitude by 2/3° longitude, and 72 vertical layers from the 122   surface to 0.01 hPa. The lowest 5 layers are centered at 70, 200, 330, 470, and 600 m for 123   a column based at sea level. We use a nested version of GEOS-Chem (Y. Wang et al.,   4 124   2004; Chen et al., 2009) with the native 1/2° × 2/3° horizontal resolution over North 125   America and adjacent oceans (140°-40°W, 10°-70°N), and 2° × 2.5° horizontal resolution 126   for the rest of the world. Zhang et al. (2011) previously used the exact same model to 127   estimate policy-relevant background ozone in surface air over the US. 128   129   The model includes a detailed simulation of tropospheric ozone-NO -hydrocarbon- x 130   aerosol chemistry, as recently described for example by Mao et al. (2010). Formation of 131   organic nitrates from the oxidation of biogenic isoprene emitted by vegetation can be a 132   significant sink for NO in the model (Horowitz et al., 1998). We assume that these x 133   isoprene nitrates are removed by wet and dry deposition at the same deposition velocity 134   as HNO and do not regenerate NO . Earlier versions of GEOS-Chem did not explicitly 3 x 135   describe isoprene nitrates, treating them instead as HNO (Bey et al., 2001). Here we 3 136   describe them explicitly in order to compare simulated HNO with observations and to 3 137   quantify the contribution of isoprene nitrates to dry deposition. 138   139   Aerosol and gas-phase chemistry in GEOS-Chem are coupled through gas-aerosol 140   partitioning of semi-volatile species including NH and HNO , heterogeneous aerosol 3 3 141   chemistry parameterized as reactive uptake coefficients (Jacob, 2000), and aerosol effects 142   on photolysis rates (Martin et al., 2003). Partitioning of total NH and HNO between the 3 3 143   gas and aerosol phases is calculated using the RPMARES thermodynamic equilibrium 144   model (Binkowski and Roselle, 2003). The reactive uptake coefficients γ for N O in N2O5 2 5 145   aerosols are from Evans and Jacob (2005), reduced by a factor of 10 as discussed in 146   Macintyre and Evans (2010). The resulting annual mean value of γ in surface air over N2O5 147   the contiguous US is 0.003, comparable to measured values in the range of 0.0005-0.006 148   (Brown et al., 2009; Bertram et al., 2009). 149   150   We conduct three-year GEOS-Chem simulations for 2006-2008. We first conduct the 151   global GEOS-Chem simulation at 2° × 2.5° resolution, and then use the output archived 152   at 3-hour temporal resolution as dynamic boundary conditions for the nested model. 153   Output from the nested model does not affect the global simulation (one-way nesting). 154     5 155   2.2. Deposition 156   157   The wet deposition scheme for aerosols is described by Liu et al. (2001), and its 158   adaptation to soluble gases follows Mari et al. (2000). It includes scavenging in 159   convective updrafts as well as in-cloud and below-cloud scavenging from large-scale 160   precipitation. In warm (liquid) clouds with T > 268 K, aerosols are assumed to be 100% 161   incorporated in cloud droplets and gases are partitioned following Henry’s law. In mixed 162   (liquid/ice) clouds (248 < T < 268 K), where precipitation takes place by riming, aerosols 163   are retained in the rime ice while gases are retained with varying efficiencies (unity for 164   HNO but 0.05 for NH ; J. Wang et al. (2008)). In cold (ice) clouds (T < 248 K), both 3 3 165   aerosols and HNO are scavenged with 100% efficiency (HNO is taken up as a 3 3 166   monolayer; Abbatt (1997)), while NH is not scavenged. 3 167   168   Dry deposition of gases and aerosols is simulated with a standard big-leaf resistance-in- 169   series model (Wesely, 1989). The dry deposition flux F out of the lowest model layer d 170   (midpoint z ≈ 70 m above the surface) is calculated as: 1 171   F = n C(z )v (z ) (1) d a 1 d 1 172   where n (molecules cm-3) is the number density of air, C(z )is the mixing ratio of the a 1 173   depositing species at height z , and v is its deposition velocity (cm s-1) at that height. The 1 d 174   deposition velocity is a function of the local meteorological condition and surface type, as 175   given by: 176   v (z )=(R (z ,z )+R +R )−1 (2) d 1 a 1 o b c 177   HereR (z ,z ) is the aerodynamic resistance to turbulent transfer from z to the 1 a 1 o 178   roughness height z close to the surface where turbulence vanishes, R is the boundary 0 b 179   layer resistance to molecular diffusion from z to the actual surface, and R is the canopy 0 c 180   or surface uptake resistance. 181   182   Table 1 lists the annual mean daytime (10-16 local time) dry deposition velocities for 183   different species computed in the model over the contiguous US. Values average 2.7 ± 184   1.5 cm s-1 for HNO , N O , and isoprene nitrates; 0.65 ± 0.40 cm s-1 for NH ; and 0.15- 3 2 5 3   6 185   0.36 cm s-1 for aerosols, NO , peroxyacetyl nitrate (PAN), and other organic nitrates. 2 186   Other nitrogen species are not significantly removed by dry deposition. Model values are 187   consistent with experimental studies, which report daytime dry deposition velocities to 188   land in the 2-10 cm s-1 range for HNO (Sievering et al., 2001; Horii et al., 2005), and in 3 189   the 0.1-1.0 cm s-1 range for PAN (Doskey et al., 2004; Turnipseed et al. 2006; Wolfe et 190   al., 2009). Biosphere-atmosphere exchange of NO and NH is bi-directional (Sutton et x 3 191   al., 1998; Lerdau et al., 2000; Ellis et al., 2011), but is treated here as uncoupled emission 192   and deposition processes. 193   194   2.3. Emissions 195   196   US anthropogenic emissions are from the EPA National Emission Inventory for 2005 197   (NEI 05) with modifications for NH described below. Anthropogenic emissions of NO , 3 x 198   CO, volatile organic compounds (VOCs), and SO outside the US are from the Emission 2 199   Database for Global Atmospheric Research (EDGAR) inventory (Olivier and Berdowski, 200   2001). Anthropogenic emissions of NH outside the US are from the Global Emission 3 201   Inventory Activity (GEIA) (Bouwman et al., 1997). These global inventories are 202   superseded by regional emission inventories from Q. Zhang et al. (2009) for Asia in 203   2006, the European Monitoring and Evaluation Program (EMEP) for Europe (Vestreng 204   and Klein, 2002), the Criteria Air Contaminants (CAC) emission inventory for Canada 205   (http://www.ec.gc.ca/pdb/cac/cac_home_e.cfm), and the Big Bend Regional Aerosol and 206   Visibility Observational (BRAVO) emission inventory for Mexico (Kuhns et al., 2005). 207   The EDGAR, EMEP, CAC, and BRAVO emissions are scaled on the basis of energy 208   statistics to 2006 as described by van Donkelaar et al. (2008). Global anthropogenic NO x 209   emissions also include fertilizer use from Yienger and Levy (1995) and aircraft from 210   Baughcum et al. (1996). 211   212   Natural NO emissions include open fires, lightning, and soil. We use monthly biomass x 213   burning emissions from the Global Fire Emission Database version 2 (GFED-v2) (van 214   der Werf et al., 2006). Lightning NO emissions are linked to deep convection following x 215   the parameterization of Price and Rind (1992) with vertical profiles from Pickering et al.   7 216   (1998). The global spatial distribution of lightning flashes is rescaled to match the 10- 217   year climatology of OTD/LIS satellite observations (Sauvage et al., 2007) with higher 218   NO yield per flash at northern mid-latitudes than in the tropics (Hudman et al., 2007). x 219   The global lightning source is imposed to be 6 Tg N a-1 (Martin et al., 2007). Soil NO x 220   emissions are computed using a modified version of the Yienger and Levy (1995) 221   algorithm with canopy reduction factors described in Wang et al. (1998). Biogenic VOC 222   emissions (important for the conversion of NO to organic nitrates) are from the Model of x 223   Emissions of Gases and Aerosols from Nature (MEGAN) (Guenther et al., 2006). Natural 224   NH emissions from soils, vegetation, and the oceans are from the GEIA inventory 3 225   (Bouwman et al., 1997). 226   227   Figure 1 shows the spatial and seasonal distribution of US NO emissions and Table 2 x 228   gives annual totals from each source over the contiguous US. Anthropogenic sources (5.6 229   Tg N a-1 including fertilizer use and aircraft) account for 84% of the total NO emissions. x 230   Natural sources from lightning, soil, and open fires account for 9.5%, 6.2%, and 0.7%, 231   respectively. Natural contributions peak in summer, accounting for 39% of US NO x 232   emissions in July. 233   234   Gilliland et al. (2003, 2006) and Pinder et al. (2006) previously found large seasonally 235   varying errors in the US NEI emission inventory for NH by model comparison with 3 236   observed wet deposition fluxes of ammonium (NH +) and atmospheric concentrations of 4 237   total reduced nitrogen (NH ≡ NH gas + ammonium aerosol). Here we use NH x 3 x 238   measurements from two networks (Figure 2, left panel) to constrain the seasonality of 239   NH emissions: the Midwest Ammonia Monitoring Project managed by the Midwest 3 240   Regional Planning Organization (RPO) for 2004-2005 (http://www.ladco.org), and the 241   Southeastern Aerosol Research and Characterization (SEARCH) for 2006 (Edgerton et 242   al., 2006). Figure 2 (central panel) compares observations to model results in a simulation 243   with the August NEI-2005 NH emission applied to the whole year (aseasonal source). 3 244   Results agree well with measurements in summer but are far too high in winter, reflecting 245   at least in part a temperature dependence of NH emission (Aneja et al., 2000). We fit 3 246   monthly scaling factors as observed/simulated concentration ratios to the NEI emissions   8 247   to correct the discrepancy shown in the central panel of Figure 2 and apply them 248   nationwide. These scaling factors range from 0.9-1 in summer to 0.2-0.4 in winter. 249   Independent comparison of the resulting model to the SEARCH data for 2006 (Figure 2, 250   right panel) shows good agreement and thus supports these seasonal scaling factors. 251   252   Figure 1 shows the spatial and seasonal distribution of US NH emissions with the above 3 253   scaling factors applied, and Table 2 gives annual totals for each source. Emissions show a 254   broad May-September maximum. The highest emissions are in areas of major livestock 255   operations. Anthropogenic emissions (2.3 Tg N a-1, 81%) dominate over natural 256   emissions (0.56 Tg N a-1, 19%). 257   258   Our NO and NH emission estimates can be compared with those of Smith et al. (2010), x 3 259   who implemented natural emission inventories into the CMAQ regional model for July 260   2002. For a model domain (130°-70°W, 23°-56°N) covering the contiguous US and large 261   fractions of Canada and Mexico, they found natural emissions to contribute 44% of NO x 262   emissions and 28% of NH emissions. Our results for July 2006 over the same CMAQ 3 263   domain are comparable, with natural emission contributions of 40% for NO and 24% for x 264   NH . The largest difference is for open fire NO emissions, which are a factor of 5 higher 3 x 265   in Smith et al. (2010). The GFED-v2 fire emissions used in GEOS-Chem indicate that 266   dry mass burned over the CMAQ domain was a factor of 4 higher in July 2002 than in 267   July 2006, mostly due to large fires in eastern Canada in 2002. 268   269   3. Deposition patterns and surface concentrations 270   271   Figure 3a-c compares simulated and observed sulfate, ammonium, and nitrate wet 272   deposition fluxes over the US and Canada for 2006. The observations are from the 251 273   sites of the National Atmospheric Deposition Program/National Trends Network 274   (NADP/NTN; data available at https://nadp.isws.illinois.edu/) for the US, and 26 sites of 275   the Canadian Air and Precipitation Monitoring Network (CAPMoN; data available at 276   http://www.on.ec.gc.ca/natchem/index.aspx) for Canada. We use sulfate as a check on the 277   wet deposition processes in the model since the SO source from coal combustion is well 2   9 278   constrained by stack measurements. We summarize the comparison between the model 279   (M) and observations (O) using the correlation coefficient, the normalized mean bias N ∑ (M −O ) 280   (NMB) computed as NMB= i=1 i i for the N CASTNet sites, and the mean N ∑ O i=1 i 1 N 281   normalized bias MNB= ∑ (M −O )/O . The NMB estimates the mean offset N i=1 i i i 282   between the model and observations, while MNB provides a sensitive evaluation of the 283   model performance for observed low values. 284   285   For all three species the model shows strong correlations with observations, no 286   significant annual biases, and little seasonal bias. MNB and NMB generally agree within a 287   few percent, except for some seasonal cases where MNB is ~20% higher due to model 288   overestimates of very low observations. Wet deposition generally peaks in summer for all 289   three species, in the case of sulfate and nitrate because of higher SO and NO oxidant 2 x 290   concentrations, and in the case of ammonium because of higher ammonia emissions. The 291   summer peak of nitrate wet deposition is particularly pronounced around the Gulf of 292   Mexico where it reflects the seasonal maxima in both lightning emissions and 293   precipitation. Simulated nitrate wet deposition also peaks in Michigan and southeastern 294   Canada in winter, caused by transport of HNO and nitrate aerosols produced mostly 3 295   from N O hydrolysis. This winter maximum is less pronounced in the observations, 2 5 296   suggesting that N O hydrolysis in the model may be too fast as discussed further below. 2 5 297   The model does not capture the observed high values of ammonium wet deposition in the 298   upper Midwest, as previously noted by Fisher et al. (2011), likely because of regional 299   underestimate of emissions. 300   301   No routine direct measurements of dry deposition fluxes are available at US sites. 302   However, the Clean Air Status and Trends Network (CASTNet) makes weekly integrated 303   measurements of gas-phase HNO concentrations from which dry deposition fluxes can 3 304   be estimated using modeled dry deposition velocities (Clarke et al., 1997). Figure 4 305   compares annual mean HNO concentrations from CASTNet with GEOS-Chem results in 3 306   2006. The model has a mean positive bias of 69%, which is due in part to the vertical   10