modeling wet deposition and concentration of inorganics over

Geosci Model Dev 5 1363ndash1375 2012wwwgeosci-model-devnet513632012doi105194gmd-5-1363-2012copy Author(s) 2012 CC Attribution 30 License

GeoscientificModel Development

Modeling wet deposition and concentration of inorganics overNortheast Asia with MRI-PMc

M Kajino 12 M Deushi1 T Maki 1 N Oshima1 Y Inomata3 K Sato3 T Ohizumi3 and H Ueda4

1Meteorological Research Institute Japan Meteorological Agency 1-1 Nagamine Tsukuba 305-0052 Japan2Pacific Northwest National Laboratory PO Box 999 Richland WA 99352 USA3Asia Center for Air Pollution Research 1182 Sowa Nishi Niigata 950-2144 Japan4Toyohashi Institute of Technology 1-1 Hibarigaoka Tempaku Toyohashi 441-8580 Japan

Correspondence toM Kajino (kajinomri-jmagojp)

Received 15 May 2012 ndash Published in Geosci Model Dev Discuss 1 June 2012Revised 9 October 2012 ndash Accepted 9 October 2012 ndash Published 6 November 2012

Abstract We conducted a regional-scale simulation overNortheast Asia for the year 2006 using an aerosol chem-ical transport model with time-varying lateral and upperboundary concentrations of gaseous species predicted bya global stratospheric and tropospheric chemistry-climatemodel The present one-way nested global-through-regional-scale model is named the Meteorological Research InstitutendashPassive-tracers Model system for atmospheric Chemistry(MRI-PMc) We evaluated the modelrsquos performance withrespect to the major anthropogenic and natural inorganiccomponents SO2minus

4 NH+

4 NOminus

3 Na+ and Ca2+ in the airrain and snow measured at the Acid Deposition MonitoringNetwork in East Asia (EANET) stations Statistical analysisshowed that approximately 40ndash50 and 70ndash80 of simu-lated concentration and wet deposition of SO2minus

4 NH+

4 NOminus

3and Ca2+ are within factors of 2 and 5 of the observationsrespectively The prediction of the sea-salt originated compo-nent Na+ was not successful at near-coastal stations (wherethe distance from the coast ranged from 150 to 700 m) be-cause the model grid resolution (1x = 60 km) is too coarseto resolve it The simulated Na+ in precipitation was signifi-cantly underestimated by up to a factor of 30

1 Introduction

Atmospheric chemical transport models have been importantfor the analysis of the emission long-range transport trans-formation and deposition of air pollutants and climate forc-ing agents and are extensively used for scientific as well as

regulatory purposes Because large discrepancies remain be-tween the model results and observational data numerousongoing efforts focus on model development and applica-tion Asia is the most populated region in the world and itsassociated anthropogenic emissions are huge (Ohara et al2007 Kurokawa et al 2009 Zhang et al 2009) In additionto these anthropogenic emissions massive amounts of Asiandust particles are lofted from arid and semi-arid regions ofthe Asian continent

Recently to develop a better common understanding of theperformance and uncertainties of chemical transport mod-els in applications in East Asia the Model Intercompar-ison Study Asia Phase II (MICS-Asia II) was performed(Carmichael et al 2008) The project included nine differ-ent models and compared each model with observations andwith all of the other models to evaluate model performancefor ozone (O3) and its related chemical species (Han et al2008) for secondary inorganic components such as sulfatenitrate and ammonium (Hayami et al 2008) and for dry andwet deposition (Wang et al 2008) One of the major findingsof the project is that an ensemble mean prediction (which isa simple average of all of the model results) agrees best withobservational data However the discrepancies between ob-servational data and the results of each model are sometimesvery large especially for the amount of wet deposition ofsulfate nitrate and ammonium For example some differ-ences were one to two orders of magnitude for the monthlyvalues (Wang et al 2008) These discrepancies were likelycaused by large uncertainties in modeling wet deposition pro-cesses Several modeling studies still lack any evaluation of

Published by Copernicus Publications on behalf of the European Geosciences Union

1364 M Kajino et al Modeling wet deposition in Northeast Asia

predicted amounts of wet deposition Such an evaluation ishowever indispensable to assess the consistency in the entiremodeling system from emissions and transport to transfor-mation and deposition

To accurately simulate the fate of Asian air pollutants amodel (Regional Air Quality Model 2 RAQM2) has beendeveloped with the following components emissions of an-thropogenic trace species biomass burning biogenic andnatural (Asian dust and sea salt) aerosol emissions advec-tion and turbulent diffusion photochemistry and new parti-cle formation gas-to-particle conversion of inorganic and or-ganic compounds Brownian coagulation CCN and IN acti-vation and cloud microphysical processes grid-scale liquid-phase chemistry in hydrometeors as well as in aerosol watersubgrid-scale convection and wet scavenging and dry depo-sition of gas and particles (details are given in Kajino et al2012 and in the current paper) Kajino et al (2012) showedthat the modeled size distributions such as PM25 PM10 oftotal mass and PM1bulk ratios of chemical componentswere consistent with the observations They used the con-stant climatological values for boundary conditions and as aresult the hemispheric transport and the intrusion of strato-spheric O3 may not have been well represented In the cur-rent study we added time-varying lateral and upper bound-ary concentrations of gaseous species predicted by a globalstratospheric and tropospheric Meteorological Research In-stitute Chemistry Climate model (MRI-CCM2 Deushi andShibata 2011) with a time resolution of 1 h The present one-way nested global-through-regional-scale model is namedMRIndashPassive-tracers Model system for atmospheric Chem-istry (MRI-PMc) In Sect 2 we describe the MRI-PMcmodel and the observational data that we used Section 3 dis-cusses the model system performance and analysis whichwas evaluated using the observational data We summarizeour major findings in Sect 4

2 Description of models and monitoring data

21 General description of MRI-PMc andparameterizations used in the model system

Figure 1 illustrates the model framework Three optionsfor objective analysis data sets for the initial and bound-ary conditions for the climate and meteorological modelsinclude The US National Center for Environmental Pre-diction (NCEP) 6 h 1 times 1 final operational global anal-ysis dataset (ds0832httpdssucaredudatasetsds0832)the Japan Meteorological Agency (JMA) Climate Data As-similation System (JCDAS) 6 h 125

times 125 (httpjrakishougojpJRA-25AboutJCDASenhtml) and the JMAMeso-Regional Objective Analysis (MANAL) data sets (3 h5 kmtimes 5 km) These options are also used for the analysisnudging method Two options for regional meteorologicalmodels include the Advanced Research Weather Research

Figure 1

Framework of MRI-PM model system

NCEP-FNL ds0832 JCDAS or MANAL

Objective Analysis Data

WRF or NHMRegional Meteorological

Model

IC BC 4DDA6 hourly

MRI-CCM2Global Stratospheric amp

Tropospheric Chemistry Climate Model

IC BC 4DDA6 hourly

Top BCLateral BC

1 hourly

RAQM2Regional Aerosol

Chemical Transport Model

Output

1 hourly

Emission InventoriesREAS INTEX-B EDGAR Bond et al (2004) GFED3 MEGAN2

Process models(Dust) Han et al (2004) (Sea-salt) Clarke et al (2006)

Meteor Var1 hourly

Gas Gas amp Aerosol

In this study WRF is selected as regional model driven by NCEP-FNL MRI-CCM2 is driven by JCDAS Only gaseous species are considered

Fig 1 Framework of Meteorological Research InstitutendashPassive-tracers Model (MRI-PM)

and Forecasting (WRF) model (version 311 Skamarock etal 2008) and the JMA nonhydrostatic model (NHM Saito etal 2007) In this study we selected WRF as a regional modeldriven by NCEP ds0832 and used a global-scale strato-spheric and tropospheric chemistry-climate model (MRI-CCM2 Deushi and Shibata 2011) for a global atmosphericsimulation driven by JCDAS

Regional Air Quality Model 2 (RAQM2 Kajino et al2012) offline coupled with the regional meteorological mod-els incorporates major atmospheric chemical and dynamicprocesses such as emissions of anthropogenic trace speciesbiomass burning biogenic and natural (dust and sea salt)aerosol emissions advection and turbulent diffusion photo-chemistry and new particle formation gas-to-particle conver-sion of inorganic and organic compounds Brownian coagu-lation CCN and IN activation and cloud microphysical pro-cesses grid-scale liquid-phase chemistry in hydrometeors aswell as in aerosol water subgrid-scale convection and wetscavenging and dry deposition of gas and particles

RAQM2 used the emissions inventory from the regionalemissions inventory in Asia (REAS Ohara et al 2007)which was extended to 2005 by Kurokawa et al (2005) andwhich included NOx SO2 NH3 non-methane volatile or-ganic compounds (NMVOCs) black carbon (BC) and pri-mary organic aerosols (POA) The speciation of NMVOCwas obtained from the INTEX-B inventory (Zhang etal 2009) For data outside the Asian region we usedEDGAR32 (Oliver and Berdowski 2001) for NOx SO2NMVOC and POA EDGAR20 (Oliver et al 1999) forNH3 (as NH3 is not provided in EDGAR32) and Bondet al (2004) for BC respectively We used the Global FireEmissions Database (GFED3 Giglio et al 2010) for openbiomass burning emissions (NOx SO2 NMVOCs BC andPOA) and the Model of Emissions of Gases and Aerosolsfrom Nature (MEGAN2 Guenther et al 2006) for bio-genic emissions of isoprene and terpenes We followed Han

Geosci Model Dev 5 1363ndash1375 2012 wwwgeosci-model-devnet513632012

M Kajino et al Modeling wet deposition in Northeast Asia 1365

et al (2004) for the dust deflation process and Clarke etal (2006) for producing sea-salt (Fig 1) As the proposedsize distribution of the Clarke module is trimodal it wasmodified into a unimodal distribution while preserving thenumber and volume concentrations with a standard deviationof σ = 2 We did not consider the surf zone emission (egde Leeuw et al 2000) as the area is not resolvable for thegrid resolution of the current model setting (1x = 60 km)

To simulate processes in the evolution of aerosol mi-croscale properties such as chemical composition size dis-tribution and mixing state we developed a new triple-moment aerosol dynamics model (Kajino 2011a b Kajinoand Kondo 2011 Kajino et al 2012) RAQM2 enablesnon-equilibrium calculations of gas-to-particle mass trans-fers over a wide range of aerosol particle diameters from1 nm to supermicron particles In RAQM2 six importantparameterizations related to aerosol dynamics and wet de-position processes are implemented (1) new particle for-mation (2) activation of cloud condensation nuclei (CCN)(3) activation of ice nuclei (IN) (4) grid-scale cloud micro-physics (5) dry deposition and (6) subgrid-scale convectionand scavenging

Figure 2 schematically illustrates the categorical ap-proach to describing aerosol dynamics and cloud micro-physics processes taking the aerosol mixing state into con-sideration The model groups aerosols into four categoriesAitken mode (ATK diametersim 10 nm) accumulation mode(ACM diametersim100 nm) soot aggregates (AGR diame-ter sim100 nm) and coarse mode (COR diametersim 1 microm)The SO2 molecule is oxidized by the OH radical in the airto form H2SO4 gas Because the vapor pressure of H2SO4gas is extremely low homogeneous nucleation and con-densation onto pre-existing particles occur simultaneouslyRAQM2 can provide a rigorous solution to this competi-tive and non-equilibrium process with a short time step of1 s The newly formed particles calculated using a param-eterization of Kuang et al (2008) are distributed into theATK category (red arrow in Fig 2) and condensation oc-curs on aerosols in all pre-existing categories (blue arrows inFig 2) Intra-category and inter-category Brownian coagula-tion (pink arrows in Fig 2) were simulated with the ModalAerosol Dynamics model for multiple Modes and fractalShapes (MADMS) (Kajino 2011a b Kajino and Kondo2011) A portion of the aerosols larger than the critical di-ameter is activated as CCN (green arrows in Fig 2) Thecritical diameter of each aerosol category is parameterizedby using the dry aerosol size distribution hygroscopic massenvironmental temperature and humidity and updraft veloc-ity (Abdul-Razzak and Ghan 2000) With regard to IN acti-vation of aerosols AGR and COR aerosols are activated bycontact freezing and immersion freezing if their hydropho-bic mass (BC OA dust) is larger than their hydrophilic mass(SO2minus

4 NOminus

3 NH+

4 Clminus sea salt) using the parameteriza-tion of Lohmann and Diehl (2010) (orange arrows in Fig 2)Portions of aerosols activated as CCN and IN are transferred

Figure 2

ATK

ACM

AGR

COR

RNW

CLD

ICE

SNW

GRW

Gas

Red New particle formationBlue CondensationDissolutionEvaporation Green Aerosol activation as Cloud Condensation NucleiOrange Aerosol activation as Ice Nuclei Purple Cloud microphysics (ie Autoconversion Accretion)Pink CoagulationCollisionCoalescence

Aerosols Hydrometeors

Fig 2 Schematic illustration of gas-aerosol-cloud dynamic pro-cesses based on the categorical approach of RAQM2

to the cloud (CLD) and cloud ice (ICE) categories respec-tively After becoming CLD and ICE particles the chemicalcomponents are transferred into larger hydrometeor particlessuch as rain (RNW) snow (SNW) or graupel (GRW) by us-ing the conversion rates parameterized by Lin et al (1983)(purple arrows in Fig 2) Aerosols in CLD and ICE that donot convert to RNW SNW or GRW within a time step are de-fined to evaporate and regenerate again after the cloud micro-physics operator The chemical components in RNW SNWand GRW are assumed to reach the ground surface instanta-neously as wet depositions Resuspension due to rain evap-oration could be an important source of sub-cloud aerosols(Saide et al 2012) but the current model does not considerthis process The model also considers coagulation betweenhydrometeors and aerosols due to gravitational settling (pinkarrows in Fig 2) and dissolution of gas into hydrometeorsand aerosol water (blue arrows in Fig 2)

The model domain common to both WRF and RAQM2is illustrated in Fig 3 which also shows the locations ofthe observation sites of the Acid Deposition Monitoring Net-work in East Asia (EANET) The horizontal grid resolutionis 60 km on a Lambert conformal map projection There are28 vertical layers from the ground to 100 hPa for WRF and13 layers from the ground to 10 km for RAQM2 with theterrain following coordinates Further descriptive details ofthe WRF-RAQM2 settings are found in Kajino et al (2012)The output time interval of the WRF is 1 h and thus the

wwwgeosci-model-devnet513632012 Geosci Model Dev 5 1363ndash1375 2012


Table 1Location information for the eight EANET remote observation sites in Japan used in the study The station locations are also shownin Fig 3

Longitude Latitude Characteristics Altitude Distance from(E) (N) (m asl) coast (m)

1 Rishiri 14112prime 4507prime Island 40 7002 Tappi 14021prime 4115prime Cape 105 3603 Ogasawara 14213prime 2705prime Island 230 5004 Sado 13824prime 3814prime Island 136 1505 Oki 13311prime 3617prime Cape 90 2006 Hedo 12815prime 2652prime Cape 60 2007 Happo 13747prime 3641prime Mountain 1850 8 Yusuhara 13256prime 3322prime Inland 790

lowast More than 10 km from the coast

Figure 3

X

Y

(m)

MRI-CCM2 (T42)

WRFRAQM2 (60km)

12

3

45

6

78

Fig 3 Modeling domains showing terrestrial elevation (m) and theEANET monitoring sites (triangles 1 through 8) Locations of theEANET stations are given in Table 1

inputoutput time interval for RAQM2 is also 1 h For lateraland upper boundary concentrations for the RAQM2 simula-tion we used hourly concentrations of NOx O3 CO andvolatile organic compounds (VOCs) that were simulated bythe MRI-CCM2 with a T42 horizontal resolution (approxi-mately 300 km) The present regional model is thus consid-ered to be a one-way nested model of the global chemistrymodel (Figs 1 and 3) The MRI-CCM2 reproduced the sea-sonal variations in measured O3 concentrations at stationsnear the edges of the RAQM2 domain

22 EANET stations and monitoring data used in thestudy

We used the EANET (Acid Deposition Monitoring Net-work in East Asia) monitoring data for model evaluation(all EANET guidelines documents and manuals are avail-able athttpwwweanetccproducthtml) The EANET sta-tions in Japan monitor 1-week (at Ogasawara) or 2-week (atother sites) accumulated concentrations of gaseous species(HNO3 HCl NH3 and SO2) and aerosol components(SO2minus

4 NOminus

3 Clminus NH+

4 Na+ Mg2+ K+ and Ca2+) usingthe filter pack (FP) method The four-stage filter pack is com-posed of four filters in line with the air streams entering from

the bottom with a flow rate of 1 l minminus1 Aerosols are col-lected on the bottom (ie first) pack with Teflon filter (poresize 08 microm diameter 47 mm) to attain good efficiency incollecting sub-micron particles (EANET ldquoTechnical Docu-ments for Filter Pack Method in East Asiardquo) By compar-ing the FP method with the simultaneous measurements us-ing the Andersen sampler (a multi-stage cascade impactor)the FP method was found to efficiently collect particles upto diameters of 10 microm The stations also monitor daily accu-mulated precipitation (SO2minus

4 NOminus

3 Clminus NH+

4 Na+ Mg2+K+ and Ca2+) using wet-only precipitation collection andion chromatography analysis (EANET ldquoTechnical Manualfor Wet Deposition Monitoring in East Asiardquo) and hourlySO2 NO NOx O3 PM25 and PM10 concentrations andmeteorological parameters such as wind speed wind di-rection temperature relative humidity and solar radiationQuality assurance and quality control were conducted to en-sure that monitoring data remained of high quality in ac-cordance with EANET guidelines (EANET ldquoQuality Assur-anceQuality Control Programsrdquo)

The FP method is subject to several distinct problemswhen used for long-term sampling Because of the possibil-ity that particulate NH4NO3 and NH4Cl will volatilize whencollected on a filter during the 1-week or 2-week samplingperiod thermodynamic equilibrium may not have been at-tained In addition high humidity can reduce the concen-tration of gaseous species because the filter pack traps con-densed water To avoid these artifacts we only used totalNOminus

3 (T-NOminus

3 is [HNO3 gas plus NOminus3 aerosol]) and totalNH+

4 (T-NH+

4 is [NH3 gas plus NH+4 aerosol]) and gas-aerosol partitioning is not discussed in this study

To deduce the anthropogenic SO2minus

4 and Ca2+ originatingfrom Asian dust (calcite) we defined non-sea-salt (nss) SO2minus

4and Ca2+ to exclude the contribution of sea salt using a stan-dard mean chemical composition of seawater (DOE 1994)as follows



[nss-SO2minus

4 ] = [SO2minus

4 ] minus 0251times [Na+] (1)

[nss-Ca2+] = [Ca2+

] minus 0038times [Na+] (2)

where [ ] denotes weight concentrations in microg mminus3 To dis-tinguish the ions in precipitation from those in aerosols wedefined the precipitation ions as W-nss-SO2minus

4 W-NH+

4 W-NOminus

3 W-nss-Ca2+ and W-Na+Among EANET stations we selected eight remote stations

in Japan for model evaluation (Fig 3 Table 1) The islandstations (Rishiri Ogasawara Sado and Oki) and those onisolated capes (Tappi and Hedo) are located far from largeanthropogenic emission sources in places where there areno complex local orographically induced winds Thus airpollutant transport events at these stations mostly coincidewith synoptic-scale disturbances and are generally well re-produced by regional-scale models However because thesestations are very close (lt 1 km) to the ocean regional-scalesimulations of ocean-derived species such as sea salt oftendid not agree well with the observations Therefore we alsoincluded two inland stations (Happo and Yusuhara) in themodel evaluation although the wind fields at these locationscan be affected by the local mountainous topography andthe quantity of precipitation may not be predicted well byregional-scale models

3 Results and discussion

31 Evaluation of aerosol chemical components

The aerosol chemical components were evaluated with theFP method however because the time resolution of thismethod is biweekly (or weekly at Ogasawara) we had only24 (or 48) samples for each site This limited number ofsamples made it difficult to evaluate transport phenomenabut we were able to evaluate the quantitative consistencyof RAQM2 As explained in Sect 22 we did not use gas-aerosol partitioning of semi-volatile inorganic componentsbecause of possible artifacts however gas-aerosol partition-ing is of the utmost importance with respect to long-rangetransport because the dry and wet deposition rates of the gasand aerosol phases are very different (Kajino et al 20052008 Kajino and Ueda 2007 2011) The partitioning needsto be evaluated with other observations in the future

We separated the observed and modeled concentrations ofaerosol chemical components into near-the-coast (NC) andfar-from-the-coast (FC) groups (Fig 4) because of the dif-ferent characteristics of the site locations in the regional-scale simulation framework As we discussed in Sect 22 theNC stations are located far from large anthropogenic emis-sion sources in places where there are no complex localorographically induced winds and thus air pollutant trans-port events mostly coincide with synoptic-scale disturbances

Figure 4

(a) (b)

(c) (d)

(e) (f)

(g) (h)

(i) (j)

Aero

sol nss‐SO

42‐

Total N

H4+

Total N

O3‐

Aero

sol nss‐C

a2+

Aero

sol N

a+

NC FC

Sim (microg m

‐3)

Sim (microg m

‐3)

Sim (microg m

‐3)

Sim (microg m

‐3)

Sim (microg m

‐3)

Obs (microg m-3) Obs (microg m‐3)

Fig 4 Scatter diagrams of modeled vs measured concentrationsof biweekly aerosol chemical components nss-SO2minus

4 T-NH+

4 T-

NOminus

3 nss-Ca2+ and Na+ (top to bottom) at (left) the near-coastalstations (NC Rishiri Tappi Sado Ogasawara Oki Hedo) and(right) the stations far from the coast (FC Happo and Yusuhara)Solid lines denote the 1 1 line and dashed lines delimit the factor-of-2 envelope



Table 2Comparative statistical analysis of all observed (Obs) and simulated (Sim) data for the year 2006

Number of data Average (Obs) Average (Sim) NMBa RMSEb R2 FAC2c FAC5d

1- or 2-week average bulk concentrations (microg mminus3)

Nss-SO2minus

4 194 292 179 minus039 205 049 050 086T-NH+

4 180 099 067 minus032 055 052 066 089T-NOminus

3 186 115 290 153 261 036 040 081Nss-Ca2+ 194 014 030 124 031 050 033 069Na+ (NC)e 144 255 230 minus010 225 007 046 086Na+ (FC)f 42 025 112 342 108 027 010 052

2-week average wet deposition amount (mmol mminus2) and precipitation amount (H2O mm)

W-nss-SO2minus

4 192 082 050 minus039 075 029 043 079W-NH+

4 192 080 065 minus019 083 015 046 080W-NOminus

3 192 082 092 013 083 025 054 086W-nss-Ca2+ 192 038 020 minus046 071 016 037 071W-Na+ (NC)e 144 164 138 minus091 522 011 013 038W-Na+ (FC)f 48 101 103 0022 150 023 042 075H2O 192 734 716 minus0025 836 018 049 086

a Normalized mean bias (mean bias divided by observation average)b Root-mean-square errorc Number fraction of data within a factor of 2d Numberfraction of data within a factor of 5e Comparison of Na+ at the near-the-coast (NC) stations (Rishiri Tappi Ogasawara Sado Oki and Hedo)f Comparisonof Na+ at the far-from-the-coast (FC) stations (Happo and Yusuhara)

and are generally reproduced well by regional-scale modelsHowever because these stations are very close to the oceanwhich is the source of sea salt particles (lt 1 km Table 1)the regional-scale simulations often did not agree well withthe observations Because the selected FC stations are alsolocated far from large anthropogenic emission sources andfar from the ocean (gt 50 km) the regional-scale simulationsof both anthropogenic and oceanic components should agreewell with the observations However because wind fields canbe affected by local mountainous topography transport pat-terns may not always be reproduced well by regional-scalesimulations In the comparison between the observations andsimulations (Table 2) we separated the NC and FC groupsonly for Na+ which originated primarily from sea salt parti-cles

Results for the simulated and observed concentrations ofanthropogenic components such as nss-SO2minus

4 T-NH+

4 andT-NOminus

3 were correlated well at both the NC and FC stations(Table 2 Fig 4) The simulated averages of nss-SO2minus

4 andT-NH+

4 were approximately 30ndash40 smaller than the ob-served averages whereas the simulated average of T-NOminus

3was 25 times the observed average Because the modeledpartitioning of gases and aerosols could not be evaluated bythe measurements it was not possible to identify the reasonsfor the discrepancies with the available information Therewere positive correlations between the simulation and obser-vation (036ndash052) and the RMSEs were comparable withthe averages

We consider nss-Ca2+ to originate from Asian dust parti-cles which contain calcite Despite the large uncertainty in

simulating dust emission flux we were able to obtain a goodvalue ofR2 (050) for nss-Ca2+ Obtaining the good valuewas possible because all of the observation stations are sit-uated at a distance from the extensive source regions (egTaklamakan and Gobi deserts) and in the downwind regionwhen long-range transport of Asian dust was predominant inspring and autumn of 2006

The primary source of Na+ is sea salt particles Naturalaerosols are usually difficult to simulate because estimates oftheir emission fluxes are uncertain Whereas the averages andRMSE for Na+ at the NC stations are similar the correlationcoefficient (R2) at the NC stations is very low (007) becauseof the uncertainty in emission flux and because the stationsare very close to the source of emission ndash the ocean sur-face Generally short-term (1 h) variations of modeled Na+

concentrations are strongly correlated with observed values(such asR sim 07 in Kajino and Kondo 2011) because theemission mass flux of sea salt correlates strongly to the sur-face wind speed TheR2 of 007 is low because the results arelong-term averages (one or two weeks) of multiple stationswith variable distances from the coast (horizontal distancesfrom 150 to 700 m vertical distances from 40 to 230 m) Weobtained a largerR2 for Na+ (027) at the FC stations whichwere far from the emission source

32 Evaluation of ion concentrations in rain and snow

To evaluate the ion concentrations in precipitation we se-lected the western island station Oki (Fig 5) and the westerninland station Yusuhara (Fig 6) to illustrate differences in



Figure 5

ObservationRAQM2

(a)

(b)

(c)

(d)

(f)

(g)

(h)

(i)

(j)

(e) (k)

Fig 5 Biweekly mean of daily variations of the observed (blue)and simulated (red) wet deposition at a station near the coastOki (a) nss-SO2minus

4 (b) NH+

4 (c) NOminus

3 (d) nss-Ca2+ (e) Na+and (f) precipitation Biweekly variations of the observed (blue)and simulated (red) aerosol components(g) nss-SO2minus

4 (h) NH+

4

(i) NOminus

3 (j) nss-Ca2+ and(k) Na+

model performance for NC and FC stations We made com-parisons to the statistical results at the eight remote EANETstations (Table 2 and Fig 7)

Both time variations and values of precipitation at Oki(Station 5) were predicted well by the model for examplethe ratio of 2-week mean simulationobservation (Sim Obsratio) was 136 andR = 072 (Fig 7i j) The quantitiesof the chemical components associated with wet depositionmost of them submicron particles such as nss-SO2minus

4 andNH+

4 were also predicted well at Oki (Fig 5a b) NOminus

3 ispartly attributed to submicron particles forming NH4NO3while some percentage of NOminus3 also consisted of coarse(simseveral microm diameter) sea salt particles forming NaNO3(Kajino and Kondo 2011 Kajino et al 2012) As Henryrsquoslaw constant of HNO3 gas is high the gas-phase fractions

Figure 6

ObservationRAQM2

(a)

(b)

(c)

(d)

(f)

(g)

(h)

(i)

(j)

(e) (k)

Fig 6Same as Fig 5 but at a station far from the coast Yusuhara

can affect the wet deposition of T-NOminus3 too The time varia-tions and values of modeled NOminus

3 in precipitation also agreedwell with the observations (Fig 5c)

We found the simulated wet deposition quantity of nss-Ca2+ to be reasonable (Fig 5d) whereas that of Na+ wasgreatly underestimated (Fig 5e) The Sim Obs ratios for W-Na+ were much less than 1 at every island and cape sta-tion (Fig 7h Stations 1ndash6) and as low as 003 at Station 3whereas the same ratios were much closer to 1 at inland sta-tions (Fig 6d and Fig 7h Stations 7ndash8)

Time variations in the precipitation at Yusuhara (Station 8)were predicted well by the model but the amounts were un-derestimated (Sim Obs= 036 R = 069 Fig 7i j) Thisoutcome possibly resulted from an underestimation of thewet deposition of chemical components although the modelreproduced time variations well (R gt 07) except for Na+

(R = 047)Rain accounts for the precipitation during the summer

and throughout the year in western and southern Japanwhereas during the winter snow accounts for most precipi-tation in northern and mountainous regions Monthly mean



Figure 7

Stations

[H2O]

[Na+]

[W-Na+]

[W-NO3-]

[T-NO3-]

[NOx]

[W-NH4+]

[T-NH4+]

[W-nss-SO42-]

[nss-SO42-]

[SO2](a)

(c)

(e)

(g)

(i)

(b)

(d)

(f)

(h)

(j)

NC FCStations

NC FC

Fig 7 Comparison of statistical results at the eight EANET sta-tions (1 Rishiri 2 Tappi 3 Ogasawara 4 Sado 5 Oki 6 Hedo7 Happo and 8 Yusuhara)R and Sim Obs for gas aerosol andprecipitation of anthropogenic sulfur oxides (nss-Sa b) reducednitrogen (Red Nc d) oxidized nitrogen (Oxid Ne f) sodium(Na g h) and amounts of precipitation (H2O i j ) Dashed linesindicate (left)R of 05 and (right) the factor-of-2 envelope

temperatures drop below 0C during the winter at RishiriTappi and Happo and the monthly mean daily minimumtemperature drops below 0C at Sado Oki and Yusuharaas well Although the surface temperature may be above0C and the surface precipitation may consist entirely of liq-uid water the cloud above often harbors cold microphysi-cal processes when the upper atmospheric temperature is be-low 0C Because there were no large and systematic sea-sonal differences in the amount and ion concentrations ofpredicted precipitation (not shown) the formulation of wetdepositional processes by RAQM2 described in Fig 2 wasconsistent for both warm and cold microphysical processes

Overall 50 and 80 of the modeled and observed two-week average values of the wet deposition of anthropogeniccomponents such as nss-SO2minus

4 NH+

4 and NOminus

3 were within

factors of two and five respectively This model performanceis noteworthy because there have been discrepancies of 1 to2 orders of magnitude in several regional-scale models evenfor monthly mean values (Wang et al 2008) because model-ing the complexities of wet deposition processes is difficultThe predictability of natural components was worse than itwas for anthropogenic components because the fluxes of thenatural emissions were heterogeneous but the performancewas good 40 and 70 of the data were within factors oftwo and five respectively for the natural components Thepercentages for anthropogenic components are similar to thecorresponding percentages for the prediction of precipitationamounts by the regional-scale meteorological model 49 and 86 were within factors of two and five respectively

33 Evaluation of the model process using theconcentration and wet deposition of eachcomponent

Regression analysis of anthropogenic sulfur oxides showedsimilar results for nss-SO2minus

4 aerosols and W-nss-SO2minus

4 inprecipitation while a somewhat lower R was obtained forSO2 gas (Fig 7a) The SimObs ratio was greater for SO2than it was for the other two phases (Fig 7b) In addition touncertainty in SO2 emission amount overestimation of SO2is likely due to underestimation of dry deposition velocitiesor underestimation of the oxidation rate from S(IV) to S(VI)Because the dry deposition velocity of SO2 is moderatelyfast its underestimation may result in overestimation of thesurface concentration We formulated oxidation of SO2 in thegas phase and in the aqueous phase of cloud rain and aerosolwater droplets (Kajino et al 2012) SO2 oxidation can alsooccur as a heterogeneous oxidation on dust particle surfaces(Tang et al 2004) but we did not implement this pathway inthe model In addition we did not consider dimethyl sulfidechemistry and SOx from outside the domain boundary in thesimulation because we assumed that these components werenot substantial for the targeted region Neglect of these pro-cesses might account for the discrepancies For sulfur oxidesapproximately 50 and 80 of the data were within factorsof two and five respectively at the stations

The mean ratio of SimObs for reduced nitrogen (NHx)

was similar for the air concentration and wet depositionOf all the chemical components assessed in the studymodel performance was the best for T-NH+

4 (R2 FAC2FAC5= 052 066 089 respectively) The performance wasalso good for [W-NH+4 ] at those stations (Stations 5ndash8) wheretheR2 of precipitation was good (gt 05)

Regarding the simulated oxidized nitrogen compounds(NOx T-NOminus

3 and W-NOminus

3 ) T-NOminus

3 was overestimatedwhereas W-NOminus3 values were reasonable The overestima-tion of T-NOminus

3 was most likely due to overestimation of pho-tochemical production or underestimation of the dry and wetdeposition rates of T-NOminus3 As discussed above the dry de-position velocity of highly reactive HNO3 gas is fast and can



be two orders of magnitude faster than that of the aerosolphase Hence discrepancies in the simulated deposition ve-locities or gas-aerosol partitioning will cause discrepanciesin surface concentrations The model performance for oxi-dized nitrogen was also good and approximately 40 and80 of these data were within factors of two and five re-spectively of the observations at all of the stations Evalua-tion of the model performance for each form of T-NOminus

3 willbe required in terms of gas-aerosol partitioning (HNO3 andNOminus

3 ) and observations of mass size distribution (for sub-micron NH4NO3 and super-micron NaNO3) to identify thereasons for the discrepancies in surface concentrations andto improve the model performance for T-NOminus

3

34 A hypothesis possible influence of large sea saltparticles (LSPs) on wet deposition near-the-coast(NC) stations

Notably the simulated W-Na+ deposition at the NC stations(Fig 7h) was greatly underestimated (up to a factor of 30 atStation 3) compared to the observations whereas the simu-lated and observed depositions were closer at the FC stationsThe results in Fig 7h raise questions regarding whether un-derestimation of the wet scavenging rate of Na+ may be thesole reason for the discrepancy because Na+ concentrationsat the FC stations were overestimated However this over-estimation may not be the sole cause First the amount ofwet deposition is not always related to the surface concentra-tion because the precipitation process involves vertical trans-port and mixing of air pollutants and is therefore affectedby both upper air and near-surface constituents Second theRMSEs were always smaller or comparable to modeled orobserved averages except for the wet deposition of Na+ atthe NC stations (522 mmol mminus2) which was significantlylarger than the observed and modeled averages (164 and138 mmol mminus2 respectively) Nevertheless it is not possi-ble to explain the discrepancy because the current modelconfiguration is not appropriate to simulate it the horizon-tal grid resolution is too coarse to resolve the NC stationsand the sea salt emission in the surf zone is not consideredThe current off-line coupling framework together with thecoarse grid resolution is not suitable for resolving the cloudmicrophysics process or for tracking the aerosols involved init

Another reason may explain the large underestimation inthe Na+ wet deposition We presumed here that large sea saltparticles (LSPs diametergt 10 microm) contributed substantiallyto the large values of [W-Na+] at the NC stations Whilethe wet-only precipitation sampler can collect any size of thedroplets during precipitation events the aerosol sampler can-not We stated that the FP method efficiently collected parti-cles of up to diameters of 10 microm but the collection efficiencycould be substantially lower for particles larger than 10 micromThis lower efficiency is due to the flow rate being too low(1 l minminus1) for LSPs which have a large downward velocity

to reach the bottom of the pack (see configuration of the in-struments in EANET ldquoTechnical Documents for Filter PackMethod in East Asiardquo athttpwwweanetccproducthtml)With a flow rate of 1 l minminus1 and with a filter diameter of47 mm the average flow rate is approximately 02 cm sminus1

(= vπd2d = 47 mmv = 1 l minminus1) This speed is the case

for pack without the filter thus the actual flow rate shouldbe lower than this The gravitational settling velocity of par-ticles with D = 10 microm is approximately 1 cm sminus1 thus thecollection efficiency of particles larger than 10 microm could belower for the current settings This result is consistent withthe fact that modeled aerosol Na+ agreed quantitatively withthe observations at NC stations

Coarse-mode particles (diametersim several microm) are pro-duced by bubbles bursting within whitecaps (an indirect pro-cess) whereas LSPs are produced when strong winds disruptwave crests and tear off drops of spumes (a direct process)(Monahan 1971) The diameter of sea salt particles producedvia the direct process could be larger than 100 microm The sur-face wind speed at 10 m elevation needs to begt 10 m sminus1

for winds to tear off the crests over the open ocean Someof the NC stations were located on rocky cliffs by the oceanWaves hitting the cliffs also produced substantial amountsof LSPs and may have contributed to the effects LSPs arealso produced under lower wind speed cases in the surf zoneand the amounts could be substantially higher than those inthe open ocean (de Leeuw et al 2000) Because such LSPsare not long lived in the air they are not transported overlong distances thus the direct process is not usually consid-ered in regional- and global-scale transport models (Gong etal 1997) Because the dry deposition velocities of LSPs andtheir collection efficiency by hydrometeors are high LSPsare efficiently removed from the air and collected in the wet-only deposition samples

Suto et al (2010) estimated that depositional mass fluxesof sea salt particles of whole size ranges (coarse mode plusLSPs) decreased by one order of magnitude at distances ofseveral kilometers from Tokyo Bay and by two orders ofmagnitude at distances of several tens of kilometers from thecoast LSPs can be transported approximately 1ndash10 km hori-zontally from the coast indicating that the horizontal repre-sentation of W-Na+ and W-Clminus the wet deposition compo-nents derived from sea salt is limited by the traveling dis-tance of LSPs

More than half of NOminus3 (and evengt 90 at Hedo) isthought to be internally mixed with coarse-mode sea salt toform NaNO3 at Japanese EANET stations located over theocean (Kajino and Kondo 2011 Kajino et al 2012) If thescavenging rate of coarse-mode sea salt in the model was un-derestimated then [W-NOminus3 ] would also be underestimatedHowever the differences between [W-NOminus

3 ] and [T-NOminus

3 ]are much smaller than those between [W-Na+] and [Na+]which is one order of magnitude or more (Fig 7f and h)Because LSPs are short lived they remain pure and are not



mixed with NOminus

3 Therefore our presumption is consistentwith the result that only the modeled [W-Na+] at NC stationswas significantly underestimated The wet deposition pro-cesses are not the same between T-NOminus

3 and Na+ becauseHNO3 gas dissolved in precipitation may also contribute to[W-NOminus

3 ] Nevertheless the feature is the same when theconcentration of HNO3 gas was extremely low under coldwinter temperatures

35 Estimating changes in pH of precipitation inducedby large sea salt particles (LSPs)

EANET has been monitoring the pH of precipitation in Asiasince 2001 There are currently 12 stations in Japan and halfof them are situated very close to the coast The pH of seawater is approximately 8 which is much higher than that ofprecipitation in Japan (pH 4sim 5) If our hypothesis presumedin Sect 34 is true the short-lived LSPs could raise the pHvalues of precipitation in Japan and those values were notrepresentative of precipitation for all of Japan (representingonly the traveling distance of LSPs 1ndash10 km) However be-cause the salinity of precipitation is much lower than thatof sea water (approximately three orders of magnitude esti-mated from the averages of Table 2) the pH of the dilutedseawater becomes almost 7 and thus should not affect thepH of precipitation substantially We proved this fact fromthe following calculation

We used the method of Walcek and Taylor (1986) to es-timate the effects of LSPs on precipitation pH by iterativelysolving the following charge-balance equation for the con-centrations of positive and negative ions

[H+] + [NH+

4 ] + 2([Ca2+] + [Mg2+

]) + [Na+] + [K+

]

= [OHminus] + [HSOminus

3 ] (3)

+ [Clminus] + [HCOminus

3 ] + 2([SO2minus

4 ]

+ [CO2minus

3 ] + [SO2minus

3 ]) + [NOminus

3 ] + [HSOminus

4 ]

We assumed the chemical components of sea water fol-lowing Song and Carmichael (2001) CO2 mixing ratio at360 ppm ocean surface salinity of 35 and obtained a consis-tent value of sea water pH (82) To assess the applicability ofthe method we compared the simulated pH of daily precip-itation using the measured chemical compositions ([SO2minus

4 ][NOminus

3 ] [Clminus] [NH+

4 ] [Ca2+] [Mg2+] [Na+] and [K+])with the measured pH at the EANET NC stations (Stations 1ndash6) For the simulation we assumed the initial gas phase con-centration of HNO3 HCl and NH3 to be equilibrated withthe measured [NOminus3 ] [Clminus] and [NH+

4 ] at the measured pHThe simulated median pH (494) was slightly higher than theobserved pH (482) but we obtained a highR2 (083) andlow RMSE (037) The slope of a regression line that fit theobserved and simulated pH values (Fig 8a) and had zerointercept was 1042

Based on the previous discussion and together withFig 7h it seems reasonable to assume that LSPs accounted

Figure 8

Obs

Sim

Obs Obs

(a) (b) (c)

Precipitation pHPrecipitation pH

(SLSD=90)

Precipitation pH(SLSD=99)

Fig 8 Scatter diagrams of daily precipitation pH at NC stations(x-axis) observed and (y-axis) calculated using(a) observed chem-ical compositions (blue)(b) with 90 of the components derivedfrom sea salt excluded (green) and(c) with 99 of the componentsderived from sea salt excluded (pink)

for 90 of the total sea salt components in precipitation atthe NC stations (distancelt 1 km) The calculated pH of theprecipitation decreased slightly when we excluded LSP ef-fects (green points in Fig 8b) Table 3 summarizes the av-erage 75th percentile median and 25th percentile values ofobserved and simulated daily precipitation pH at the NC sta-tions The average pH was not an arithmetic average of dailypH values but instead was equated to minus the commonlogarithm of the average of the daily [H+] concentrationsThe simulated effect of 90 LSP contamination of the seasalt resulted in an increase in pH by 0014 from 469 to 471on average This increment is very small and indeed muchsmaller than the RMSE (037) between the simulated andobserved pH Even if we assumed that LSPs accounted for99 of the total sea salt components in precipitation the ef-fect on pH was only approximately 01 (Fig 8c Table 3)Therefore the proximity to the coast of the remote EANETstations on islands or isolated capes does not seriously con-found the monitoring of background levels of precipitationpH Nevertheless careful attention is needed as the substan-tial contribution of LSPs may prohibit accurate determina-tion of [W-nss-SO2minus

4 ] or [W-nss-Ca2+] in precipitation bysubtracting the huge [W-Na+] term in Eqs (1) and (2)

4 Conclusions

We conducted a regional-scale simulation using an aerosolchemical transport model (RAQM2 Kajino et al 2012) withtime-varying lateral and upper boundary concentrations ofgaseous species that were predicted with a global strato-spheric and tropospheric chemistry-climate model (MRI-CCM2 Deushi and Shibata 2011) The current model sys-tem is referred to as the MRIndashPassive-tracers Model systemfor atmospheric Chemistry (MRI-PMc) We demonstratedits performance with respect to major anthropogenic andnatural inorganic components SO2minus

4 NH+

4 NOminus

3 Na+ andCa2+ in the air as well as in rain and snow precipitationStatistical analysis showed that approximately 40ndash50 and



Table 3Statistical values of observed and simulated daily precipi-tation pH at all the NC stations (Stations 1ndash6)

Averagea 75 Median 25percentile percentile

Observed pH 460 518 482 449

Simulated pH 471 536 494 459

Simulated pH(LSP= 90 )b

469 536 493 457

Simulated pH(LSP= 99 )c

460 528 487 451

1pH(LSP= 90 )b

0014 0027 00096 00024

1pH(LSP= 99 )c

010 014 0033 00062

a Minus the common logarithm of the average of the daily [H+] b Simulationwith observed chemical compositions but excluding 90 of the componentsoriginated from sea saltc Simulation with observed chemical compositions butexcluding 99 of the components originated from sea salt

70ndash80 of simulated concentration and wet deposition ofSO2minus

4 NH+

4 NOminus

3 and Ca2+ are within factors of 2 and 5 ofthe observations respectively The performance in the simu-lation of the wet deposition amount is nevertheless notewor-thy because there had been differences of one to two ordersof magnitude among several regional-scale models for sim-ulating monthly mean values of the wet deposition of SO2minus

4 NH+

4 and NOminus

3 (Wang et al 2008)The prediction of the sea-salt originated component Na+

was not successful at near-coastal stations (where the dis-tance from the coast was from 150 to 700 m) because themodel grid resolution (1x = 60 km) is too coarse to resolveit We found a large underestimation of Na+ (up to a factorof 30) in precipitation The RMSEs were always smaller orcomparable to modeled or observed averages except for thewet deposition of Na+ (522 mmol mminus2) which was signifi-cantly larger than the observed and modeled averages (164and 138 mmol mminus2 respectively) We hypothesized that thisresult was most likely due to the contribution of short-livedlarge sea salt particles (LSPs diameters larger than 10 microm) inrainwater samples Because of the presence of these dropletsmeasurements of the compositions of wet deposition derivedfrom sea salt such as Na+ and Clminus would not be representa-tive beyond the horizontal traveling distance of LSPs How-ever we calculated the effects of LSPs on precipitation pH tobe very low a change of approximately +0014 on averagebased on the assumption that LSPs were the source of 90 of the components of sea salt in precipitation Even if LSPsaccounted for 99 of total sea salt the change of pH wouldbe only approximately +01 Therefore the proximity to thecoast of the remote EANET stations is not a serious problemfor monitoring background levels of precipitation pH

The discrepancy between the simulated and observed sur-face [T-NOminus

3 ] values was greater than it was for any other

components Nitrates exist in various forms in the air in-cluding HNO3 gas NH4NO3 in submicron particles andNaNO3 in coarse-mode sea salt particles Dry and wet de-position rates might be substantially different among thesethree forms To improve the simulation of [T-NOminus

3 ] we needto assess the model performance for each form of nitrate bycomparing the simulated results with measurements of gas-aerosol partitioning and of mass size distributions of nitrate

Finally online coupling of regional-scale meteorologywith a chemical transport model with finer grid resolutionswill be essential for better simulation of the processes of wetdeposition Although RAQM2 could ldquodiagnoserdquo autoconver-sion and accretion rates with a mixed-phase parameteriza-tion of cloud microphysics these rates were based on envi-ronmental and microphysical parameters with 60 km of gridspacing at a time resolution of 1 h which is too coarse andtoo long for the special and temporal scale of dynamic andmicrophysical processes in clouds

AcknowledgementsThis research was supported by the Funda-mental Research Budget of the Meteorological Research Instituteof Japan ldquoStudies on Properties and Processes of AtmosphericAerosolsrdquo The study was partly supported by the EnvironmentResearch and Technology Development Fund (Project No B-0905and A-1101) of the Ministry of the Environment of Japan andthe Ministry of Education Science Sports and Culture (MEXT)Grant-in-Aid for Scientific Research (B) 23310018 2011M K thanks Hiroshi Hara of Tokyo University of Agriculture andTechnology Kazuhide Matsuda of Meisei University and NaotoKihara of the Central Research Institute of the Electric PowerIndustry for the useful discussions

Edited by H Tost

References

Abdul-Razzak H and Ghan S J A parameterization of aerosol ac-tivation 2 Multiple aerosol typles J Geophys Res 105 6837ndash6844 2000

Bond T C Streets D G Yarber K F Nelson S M Woo JndashHand Klimont Z A technology-based global inventory of blackand organic carbon emission from combustion J Geophys Res109 D14203doi1010292003JD003697 2004

Carmichael G R Sakurai T Streets D Hozumi Y Ueda HPark S U Fung C Han Z Kajino M Engardt M BannetC Hayami H Sartelet K Holloway T Wang Z KannariA Fu J Matsuda K Thongboonchoo N and Amann MMICS-Asia II the model intercomparison study for Asia phaseII methodology and overview of findings Atmos Environ 423468ndash3490 2008

Clarke A D Owens S R and Zhou J An ultrafine sea-saltflux from breaking waves Implications for cloud condensationnuclei in the remote marine atmosphere J Geophys Res 111D06202doi1010292005JD006565 2006

de Leeuw G Neele F P Hill M Smith M H and VignatiE Production of sea spray aerosol in the surf zone J GeophysRes 105 29397ndash29409 2000



Deushi M and Shibata K Development of an MRI Chemistry-Climate Model ver2 for the study of tropospheric and straro-spheric chemistry Papers in Meteor Geophys 62 1ndash46 2011

DOE 1994 Handbook of methods for the analysis of the variousparameters of the carbon dioxide system in sea water Version2 edited by Dickson A G and Goyet C ORNLCDIAC-741994

Giglio L Randerson J T van der Werf G R Kasibhatla PS Collatz G J Morton D C and DeFries R S Assess-ing variability and long-term trends in burned area by mergingmultiple satellite fire products Biogeosciences 7 1171ndash1186doi105194bg-7-1171-2010 2010

Gong S L Barrie L A and Blanchet J-P Modeling sea saltaerosols in the atmosphere 1 Model development J GeophysRes 102 3805ndash3818 1997

Guenther A Karl T Harley P Wiedinmyer C Palmer P Iand Geron C Estimates of global terrestrial isoprene emissionsusing MEGAN (Model of Emissions of Gases and Aerosols fromNature) Atmos Chem Phys 6 3181ndash3210doi105194acp-6-3181-2006 2006

Han Z Ueda H Matsuda K Zhang R Arao K Kanai Yand Hasome H Model study on particle size segregation anddeposition during Asian dust events in March 2002 J GeophysRes 109 D19205doi1010292004JD004920 2004

Han Z Sakurai T Ueda H Carmichael G R Streets DHayami H Wang Z Holloway T Engardt M Hozumi YPark S U Kajino M Sartelet K Fung C Bennet C Thong-boonchoo N Tang Y Chang A Matsuda K and Amann MMICS-Asia II Model intercomparison and evalution of ozoneand relevant species Atmos Environ 42 3491ndash3509 2008

Hayami H Sakurai T Han Z Ueda H Carmichael G RStreets D Holloway T Wang Z Thongboonchoo N En-gardt M Bennet C Fung C Chang A Park S U KajinoM Sartelet K Matsuda K and Amann M MICS-Asia IIModel intercomparison and evalution of particulate sulfate ni-trate and ammonium Atmos Environ 42 3510ndash3527 2008

Kajino M MADMS Modal Aerosol Dynamics model for mul-tiple Modes and fractal Shapes in the free-molecular and near-continuum regimes J Aerosol Sci 42 224ndash248 2011a

Kajino M Development of an efficient but accurate new dynamicsmodel to predict a variety of atmospheric aerosol properties andtheir elemental processes Earozoru Kenkyu 26 296ndash306 2011b(in Japanese)

Kajino M and Kondo Y EMTACS Development and regional-scale simulation of a size chemical mixing type and soot shaperesolved atmospheric particle model J Geophys Res 116D02303doi1010292010JD015030 2011

Kajino M and Ueda H Chapter 2 Secondary AcidificationMonitoring Control and Effects of Air Pollution edited byChmielewski A G In Tech Open Access Publisher ISBN978-953-307-526-6 15ndash38 2011

Kajino M Ueda H Satsumabayashi H and Han Z Increasein nitrate and chloride deposition in East Asia due to increasedsulfate associated with the eruption of Miyakejima Volcano JGeophys Res 110 D18203doi1010292005JD005879 2005

Kajino M and Ueda H Chapter 26 Increase in nitrate depositionas a result of sulfur dioxide emission increase in Asia Indirectacidification Develop Environ Sci 6 134ndash143 2007

Kajino M Ueda H and Nakayama S Secondary acidifica-tion Changes in gas-aerosol partitioning of semivolatile nitricacid and enhancement of its deposition due to increased emis-sion and concentration of SOx J Geophys Res 113 D03302doi1010292007JD008635 2008

Kajino M Inomata Y Sato K Ueda H Han Z An J KatataG Deushi M Maki T Oshima N Kurokawa J Ohara TTakami A and Hatakeyama S Development of an aerosolchemical transport model RAQM2 and predictions of NortheastAsian aerosol mass size chemistry and mixing type AtmosChem Phys Discuss 12 13405ndash13456doi105194acpd-12-13405-2012 2012

Kuang C McMurry P H McCormick A V and Eisele F LDependence of nucleation rates on sulfuric acid vapor concen-tration in diverse atmospheric locations J Geophys Res 113D10209doi1010292007JD009253 2008

Kurokawa J Ohara T Uno I Hayasaki M and Tanimoto HInfluence of meteorological variability on interannual variationsof springtime boundary layer ozone over Japan during 1981ndash2005 Atmos Chem Phys 9 6287ndash6304doi105194acp-9-6287-2009 2009

Lin YndashL Farley R D and Orville H D Bulk parameterizationof the snow field in a cloud model J Clim Appl Meteorol 221065ndash1092 1983

Lohmann U and Diehl K Sensitivity studies of the importance ofdust ice nuclei for the indirect aerosol effect on stratiform mixed-phase clouds J Atmos Sci 63 968ndash982 2010

Monahan E C Oceanic whitecaps J Phys Oceanogr 1 139ndash144 1971

Ohara T Akimoto H Kurokawa J Horii N Yamaji K YanX and Hayasaka T An Asian emission inventory of an-thropogenic emission sources for the period 1980ndash2020 At-mos Chem Phys 7 4419ndash4444doi105194acp-7-4419-20072007

Oliver J G J and Berdowski J J M Global emission sourcesand sinks in The Climate System edited by Berdowski JGuicherit R and Heij B J A A Balkema Lisse Netherlands33ndash78 2001

Oliver J G J Bouwman A F Berdowski J J M Veldt CBloos J P Jp Visschedijk A J H Van der Maas C W M andZandveld P Y J Sectoral emission inventories of greenhousegases for 1990 on a per country basis as well as on 1times 1 Envi-ron Sci Policy 2 241ndash263doi101016S1462-9011(99)00027-1 1999

Saide P E Spak S N Carmichael G R Mena-Carrasco MA Yang Q Howell S Leon D C Snider J R Bandy AR Collett J L Benedict K B de Szoeke S P Hawkins LN Allen G Crawford I Crosier J and Springston S REvaluating WRF-Chem aerosol indirect effects in Southeast Pa-cific marine stratocumulus during VOCALS-REx Atmos ChemPhys 12 3045ndash3064doi105194acp-12-3045-2012 2012

Saito K Ishida J Aranami K Hara T Segawa T Narita Mand Honda Y Nonhydrostatic atmospheric models and opera-tional development at JMA J Meteorol Soc Jpn 85B 271ndash304doi102151jmsj85B271 2007

Song C-H and Carmichael G R A three-dimensional modelinginvestigation of the evolution processes of dust and sea-salt par-ticles in east Asia J Geophys Res 106 18131ndash18154 2001



Skamarock W C Klemp J B Dudhia J Gill D O Barker DM Duda M G Huang X Y Wang W and Powers J GA description of the advanced research WRF version 3 TechNote NCARTNsim475+STR Natl Cent for Atmos Res Boul-der Colo 125 pp 2008

Suto H Hattori Y Hirakuchi H and Kihara N Study of estima-tion of salt deposition with numerical simulation of sea salt parti-cle transport (Part 4) ndash Application of estimation method for widearea distribution of sea salt to the Kanto region Civil Engineer-ing Research Laboratory Rep No N10006 CRIEPI ISBN978-4-7983-0412-0 2010 (in Japanese)

Tang Y Carmichael G R Kurata G Uno I Weber R J SongC-H Guttikunda S K Woo JndashH Streets D G Wei CClarke A D Huebert B and Anderson T L Impacts of duston regional tropospheric chemistry during the ACE-Asia exper-iment A model study with observations J Geophys Res 109D19S21doi1010292003JD003806 2004

Walcek C J and Taylor G R A theoretical method for comput-ing vertical distributions of acidity and sulfate production withincumulus clouds J Atmos Sci 43 339ndash355 1986

Wang Z Xie F Sakurai T Ueda H Han Z CarmichaelG R Streets D Engardt M Holloway T Hayami H Ka-jino M Thongboonchoo N Bennet C Park S U Fung CChang A Sartelet K Amann M MICS-Asia II Model inter-comparison and evaluation of acid deposition Atmos Environ42 3528ndash3542 2008

Zhang Q Streets D G Carmichael G R He K B Huo HKannari A Klimont Z Park I S Reddy S Fu J S ChenD Duan L Lei Y Wang L T and Yao Z L Asian emis-sions in 2006 for the NASA INTEX-B mission Atmos ChemPhys 9 5131ndash5153doi105194acp-9-5131-2009 2009



predicted amounts of wet deposition Such an evaluation ishowever indispensable to assess the consistency in the entiremodeling system from emissions and transport to transfor-mation and deposition

To accurately simulate the fate of Asian air pollutants amodel (Regional Air Quality Model 2 RAQM2) has beendeveloped with the following components emissions of an-thropogenic trace species biomass burning biogenic andnatural (Asian dust and sea salt) aerosol emissions advec-tion and turbulent diffusion photochemistry and new parti-cle formation gas-to-particle conversion of inorganic and or-ganic compounds Brownian coagulation CCN and IN acti-vation and cloud microphysical processes grid-scale liquid-phase chemistry in hydrometeors as well as in aerosol watersubgrid-scale convection and wet scavenging and dry depo-sition of gas and particles (details are given in Kajino et al2012 and in the current paper) Kajino et al (2012) showedthat the modeled size distributions such as PM25 PM10 oftotal mass and PM1bulk ratios of chemical componentswere consistent with the observations They used the con-stant climatological values for boundary conditions and as aresult the hemispheric transport and the intrusion of strato-spheric O3 may not have been well represented In the cur-rent study we added time-varying lateral and upper bound-ary concentrations of gaseous species predicted by a globalstratospheric and tropospheric Meteorological Research In-stitute Chemistry Climate model (MRI-CCM2 Deushi andShibata 2011) with a time resolution of 1 h The present one-way nested global-through-regional-scale model is namedMRIndashPassive-tracers Model system for atmospheric Chem-istry (MRI-PMc) In Sect 2 we describe the MRI-PMcmodel and the observational data that we used Section 3 dis-cusses the model system performance and analysis whichwas evaluated using the observational data We summarizeour major findings in Sect 4

2 Description of models and monitoring data

21 General description of MRI-PMc andparameterizations used in the model system

Figure 1 illustrates the model framework Three optionsfor objective analysis data sets for the initial and bound-ary conditions for the climate and meteorological modelsinclude The US National Center for Environmental Pre-diction (NCEP) 6 h 1 times 1 final operational global anal-ysis dataset (ds0832httpdssucaredudatasetsds0832)the Japan Meteorological Agency (JMA) Climate Data As-similation System (JCDAS) 6 h 125

times 125 (httpjrakishougojpJRA-25AboutJCDASenhtml) and the JMAMeso-Regional Objective Analysis (MANAL) data sets (3 h5 kmtimes 5 km) These options are also used for the analysisnudging method Two options for regional meteorologicalmodels include the Advanced Research Weather Research

Figure 1

Framework of MRI-PM model system

NCEP-FNL ds0832 JCDAS or MANAL

Objective Analysis Data

WRF or NHMRegional Meteorological

Model

IC BC 4DDA6 hourly

MRI-CCM2Global Stratospheric amp

Tropospheric Chemistry Climate Model

IC BC 4DDA6 hourly

Top BCLateral BC

1 hourly

RAQM2Regional Aerosol

Chemical Transport Model

Output

1 hourly

Emission InventoriesREAS INTEX-B EDGAR Bond et al (2004) GFED3 MEGAN2

Process models(Dust) Han et al (2004) (Sea-salt) Clarke et al (2006)

Meteor Var1 hourly

Gas Gas amp Aerosol

In this study WRF is selected as regional model driven by NCEP-FNL MRI-CCM2 is driven by JCDAS Only gaseous species are considered

Fig 1 Framework of Meteorological Research InstitutendashPassive-tracers Model (MRI-PM)

and Forecasting (WRF) model (version 311 Skamarock etal 2008) and the JMA nonhydrostatic model (NHM Saito etal 2007) In this study we selected WRF as a regional modeldriven by NCEP ds0832 and used a global-scale strato-spheric and tropospheric chemistry-climate model (MRI-CCM2 Deushi and Shibata 2011) for a global atmosphericsimulation driven by JCDAS

Regional Air Quality Model 2 (RAQM2 Kajino et al2012) offline coupled with the regional meteorological mod-els incorporates major atmospheric chemical and dynamicprocesses such as emissions of anthropogenic trace speciesbiomass burning biogenic and natural (dust and sea salt)aerosol emissions advection and turbulent diffusion photo-chemistry and new particle formation gas-to-particle conver-sion of inorganic and organic compounds Brownian coagu-lation CCN and IN activation and cloud microphysical pro-cesses grid-scale liquid-phase chemistry in hydrometeors aswell as in aerosol water subgrid-scale convection and wetscavenging and dry deposition of gas and particles

RAQM2 used the emissions inventory from the regionalemissions inventory in Asia (REAS Ohara et al 2007)which was extended to 2005 by Kurokawa et al (2005) andwhich included NOx SO2 NH3 non-methane volatile or-ganic compounds (NMVOCs) black carbon (BC) and pri-mary organic aerosols (POA) The speciation of NMVOCwas obtained from the INTEX-B inventory (Zhang etal 2009) For data outside the Asian region we usedEDGAR32 (Oliver and Berdowski 2001) for NOx SO2NMVOC and POA EDGAR20 (Oliver et al 1999) forNH3 (as NH3 is not provided in EDGAR32) and Bondet al (2004) for BC respectively We used the Global FireEmissions Database (GFED3 Giglio et al 2010) for openbiomass burning emissions (NOx SO2 NMVOCs BC andPOA) and the Model of Emissions of Gases and Aerosolsfrom Nature (MEGAN2 Guenther et al 2006) for bio-genic emissions of isoprene and terpenes We followed Han






4 NOminus

3 NH+


Figure 2

ATK

ACM

AGR

COR

RNW

CLD

ICE

SNW

GRW

Gas












Figure 3

X

Y

(m)

MRI-CCM2 (T42)

WRFRAQM2 (60km)

12

3

45

6

78





4 NOminus

3 Clminus NH+



4 NOminus

3 Clminus NH+



3 (T-NOminus


4 (T-NH+







[nss-SO2minus

4 ] = [SO2minus


[nss-Ca2+] = [Ca2+



4 W-NH+

4 W-NOminus







Figure 4

(a) (b)

(c) (d)

(e) (f)

(g) (h)

(i) (j)

Aero

sol nss‐SO

42‐

Total N

H4+

Total N

O3‐

Aero

sol nss‐C

a2+

Aero

sol N

a+

NC FC

Sim (microg m

‐3)

Sim (microg m

‐3)

Sim (microg m

‐3)

Sim (microg m

‐3)

Sim (microg m

‐3)



4 T-NH+

4 T-

NOminus







Nss-SO2minus

4 194 292 179 minus039 205 049 050 086T-NH+

4 180 099 067 minus032 055 052 066 089T-NOminus



W-nss-SO2minus

4 192 082 050 minus039 075 029 043 079W-NH+

4 192 080 065 minus019 083 015 046 080W-NOminus





4 T-NH+

4 andT-NOminus


4 andT-NH+











Figure 5

ObservationRAQM2

(a)

(b)

(c)

(d)

(f)

(g)

(h)

(i)

(j)

(e) (k)


4 (b) NH+

4 (c) NOminus


4 (h) NH+

4

(i) NOminus




4 andNH+



Figure 6

ObservationRAQM2

(a)

(b)

(c)

(d)

(f)

(g)

(h)

(i)

(j)

(e) (k)










Figure 7

Stations

[H2O]

[Na+]

[W-Na+]

[W-NO3-]

[T-NO3-]

[NOx]

[W-NH4+]

[T-NH4+]

[W-nss-SO42-]

[nss-SO42-]

[SO2](a)

(c)

(e)

(g)

(i)

(b)

(d)

(f)

(h)

(j)

NC FCStations

NC FC




4 NH+

4 and NOminus

3 were within










3 and W-NOminus

3 ) T-NOminus








3











3 ] and [T-NOminus




mixed with NOminus







[H+] + [NH+

4 ] + 2([Ca2+] + [Mg2+

]) + [Na+] + [K+

]


3 ] (3)


3 ] + 2([SO2minus

4 ]

+ [CO2minus

3 ] + [SO2minus

3 ]) + [NOminus

3 ] + [HSOminus

4 ]


4 ][NOminus

3 ] [Clminus] [NH+




Figure 8

Obs

Sim

Obs Obs

(a) (b) (c)


(SLSD=90)





4 Conclusions


4 NH+

4 NOminus






Observed pH 460 518 482 449

Simulated pH 471 536 494 459


469 536 493 457


460 528 487 451

1pH(LSP= 90 )b

0014 0027 00096 00024

1pH(LSP= 99 )c

010 014 0033 00062



4 NH+

4 NOminus


4 NH+

4 and NOminus









Edited by H Tost

References
















































4 NOminus

3 NH+


Figure 2

ATK

ACM

AGR

COR

RNW

CLD

ICE

SNW

GRW

Gas












Figure 3

X

Y

(m)

MRI-CCM2 (T42)

WRFRAQM2 (60km)

12

3

45

6

78





4 NOminus

3 Clminus NH+



4 NOminus

3 Clminus NH+



3 (T-NOminus


4 (T-NH+







[nss-SO2minus

4 ] = [SO2minus


[nss-Ca2+] = [Ca2+



4 W-NH+

4 W-NOminus







Figure 4

(a) (b)

(c) (d)

(e) (f)

(g) (h)

(i) (j)

Aero

sol nss‐SO

42‐

Total N

H4+

Total N

O3‐

Aero

sol nss‐C

a2+

Aero

sol N

a+

NC FC

Sim (microg m

‐3)

Sim (microg m

‐3)

Sim (microg m

‐3)

Sim (microg m

‐3)

Sim (microg m

‐3)



4 T-NH+

4 T-

NOminus







Nss-SO2minus

4 194 292 179 minus039 205 049 050 086T-NH+

4 180 099 067 minus032 055 052 066 089T-NOminus



W-nss-SO2minus

4 192 082 050 minus039 075 029 043 079W-NH+

4 192 080 065 minus019 083 015 046 080W-NOminus





4 T-NH+

4 andT-NOminus


4 andT-NH+











Figure 5

ObservationRAQM2

(a)

(b)

(c)

(d)

(f)

(g)

(h)

(i)

(j)

(e) (k)


4 (b) NH+

4 (c) NOminus


4 (h) NH+

4

(i) NOminus




4 andNH+



Figure 6

ObservationRAQM2

(a)

(b)

(c)

(d)

(f)

(g)

(h)

(i)

(j)

(e) (k)










Figure 7

Stations

[H2O]

[Na+]

[W-Na+]

[W-NO3-]

[T-NO3-]

[NOx]

[W-NH4+]

[T-NH4+]

[W-nss-SO42-]

[nss-SO42-]

[SO2](a)

(c)

(e)

(g)

(i)

(b)

(d)

(f)

(h)

(j)

NC FCStations

NC FC




4 NH+

4 and NOminus

3 were within










3 and W-NOminus

3 ) T-NOminus








3











3 ] and [T-NOminus




mixed with NOminus







[H+] + [NH+

4 ] + 2([Ca2+] + [Mg2+

]) + [Na+] + [K+

]


3 ] (3)


3 ] + 2([SO2minus

4 ]

+ [CO2minus

3 ] + [SO2minus

3 ]) + [NOminus

3 ] + [HSOminus

4 ]


4 ][NOminus

3 ] [Clminus] [NH+




Figure 8

Obs

Sim

Obs Obs

(a) (b) (c)


(SLSD=90)





4 Conclusions


4 NH+

4 NOminus






Observed pH 460 518 482 449

Simulated pH 471 536 494 459


469 536 493 457


460 528 487 451

1pH(LSP= 90 )b

0014 0027 00096 00024

1pH(LSP= 99 )c

010 014 0033 00062



4 NH+

4 NOminus


4 NH+

4 and NOminus









Edited by H Tost

References

















































Figure 3

X

Y

(m)

MRI-CCM2 (T42)

WRFRAQM2 (60km)

12

3

45

6

78





4 NOminus

3 Clminus NH+



4 NOminus

3 Clminus NH+



3 (T-NOminus


4 (T-NH+







[nss-SO2minus

4 ] = [SO2minus


[nss-Ca2+] = [Ca2+



4 W-NH+

4 W-NOminus







Figure 4

(a) (b)

(c) (d)

(e) (f)

(g) (h)

(i) (j)

Aero

sol nss‐SO

42‐

Total N

H4+

Total N

O3‐

Aero

sol nss‐C

a2+

Aero

sol N

a+

NC FC

Sim (microg m

‐3)

Sim (microg m

‐3)

Sim (microg m

‐3)

Sim (microg m

‐3)

Sim (microg m

‐3)



4 T-NH+

4 T-

NOminus







Nss-SO2minus

4 194 292 179 minus039 205 049 050 086T-NH+

4 180 099 067 minus032 055 052 066 089T-NOminus



W-nss-SO2minus

4 192 082 050 minus039 075 029 043 079W-NH+

4 192 080 065 minus019 083 015 046 080W-NOminus





4 T-NH+

4 andT-NOminus


4 andT-NH+











Figure 5

ObservationRAQM2

(a)

(b)

(c)

(d)

(f)

(g)

(h)

(i)

(j)

(e) (k)


4 (b) NH+

4 (c) NOminus


4 (h) NH+

4

(i) NOminus




4 andNH+



Figure 6

ObservationRAQM2

(a)

(b)

(c)

(d)

(f)

(g)

(h)

(i)

(j)

(e) (k)










Figure 7

Stations

[H2O]

[Na+]

[W-Na+]

[W-NO3-]

[T-NO3-]

[NOx]

[W-NH4+]

[T-NH4+]

[W-nss-SO42-]

[nss-SO42-]

[SO2](a)

(c)

(e)

(g)

(i)

(b)

(d)

(f)

(h)

(j)

NC FCStations

NC FC




4 NH+

4 and NOminus

3 were within










3 and W-NOminus

3 ) T-NOminus








3











3 ] and [T-NOminus




mixed with NOminus







[H+] + [NH+

4 ] + 2([Ca2+] + [Mg2+

]) + [Na+] + [K+

]


3 ] (3)


3 ] + 2([SO2minus

4 ]

+ [CO2minus

3 ] + [SO2minus

3 ]) + [NOminus

3 ] + [HSOminus

4 ]


4 ][NOminus

3 ] [Clminus] [NH+




Figure 8

Obs

Sim

Obs Obs

(a) (b) (c)


(SLSD=90)





4 Conclusions


4 NH+

4 NOminus






Observed pH 460 518 482 449

Simulated pH 471 536 494 459


469 536 493 457


460 528 487 451

1pH(LSP= 90 )b

0014 0027 00096 00024

1pH(LSP= 99 )c

010 014 0033 00062



4 NH+

4 NOminus


4 NH+

4 and NOminus









Edited by H Tost

References













































[nss-SO2minus

4 ] = [SO2minus


[nss-Ca2+] = [Ca2+



4 W-NH+

4 W-NOminus







Figure 4

(a) (b)

(c) (d)

(e) (f)

(g) (h)

(i) (j)

Aero

sol nss‐SO

42‐

Total N

H4+

Total N

O3‐

Aero

sol nss‐C

a2+

Aero

sol N

a+

NC FC

Sim (microg m

‐3)

Sim (microg m

‐3)

Sim (microg m

‐3)

Sim (microg m

‐3)

Sim (microg m

‐3)



4 T-NH+

4 T-

NOminus







Nss-SO2minus

4 194 292 179 minus039 205 049 050 086T-NH+

4 180 099 067 minus032 055 052 066 089T-NOminus



W-nss-SO2minus

4 192 082 050 minus039 075 029 043 079W-NH+

4 192 080 065 minus019 083 015 046 080W-NOminus





4 T-NH+

4 andT-NOminus


4 andT-NH+











Figure 5

ObservationRAQM2

(a)

(b)

(c)

(d)

(f)

(g)

(h)

(i)

(j)

(e) (k)


4 (b) NH+

4 (c) NOminus


4 (h) NH+

4

(i) NOminus




4 andNH+



Figure 6

ObservationRAQM2

(a)

(b)

(c)

(d)

(f)

(g)

(h)

(i)

(j)

(e) (k)










Figure 7

Stations

[H2O]

[Na+]

[W-Na+]

[W-NO3-]

[T-NO3-]

[NOx]

[W-NH4+]

[T-NH4+]

[W-nss-SO42-]

[nss-SO42-]

[SO2](a)

(c)

(e)

(g)

(i)

(b)

(d)

(f)

(h)

(j)

NC FCStations

NC FC




4 NH+

4 and NOminus

3 were within










3 and W-NOminus

3 ) T-NOminus








3











3 ] and [T-NOminus




mixed with NOminus







[H+] + [NH+

4 ] + 2([Ca2+] + [Mg2+

]) + [Na+] + [K+

]


3 ] (3)


3 ] + 2([SO2minus

4 ]

+ [CO2minus

3 ] + [SO2minus

3 ]) + [NOminus

3 ] + [HSOminus

4 ]


4 ][NOminus

3 ] [Clminus] [NH+




Figure 8

Obs

Sim

Obs Obs

(a) (b) (c)


(SLSD=90)





4 Conclusions


4 NH+

4 NOminus






Observed pH 460 518 482 449

Simulated pH 471 536 494 459


469 536 493 457


460 528 487 451

1pH(LSP= 90 )b

0014 0027 00096 00024

1pH(LSP= 99 )c

010 014 0033 00062



4 NH+

4 NOminus


4 NH+

4 and NOminus









Edited by H Tost

References
















































Nss-SO2minus

4 194 292 179 minus039 205 049 050 086T-NH+

4 180 099 067 minus032 055 052 066 089T-NOminus



W-nss-SO2minus

4 192 082 050 minus039 075 029 043 079W-NH+

4 192 080 065 minus019 083 015 046 080W-NOminus





4 T-NH+

4 andT-NOminus


4 andT-NH+











Figure 5

ObservationRAQM2

(a)

(b)

(c)

(d)

(f)

(g)

(h)

(i)

(j)

(e) (k)


4 (b) NH+

4 (c) NOminus


4 (h) NH+

4

(i) NOminus




4 andNH+



Figure 6

ObservationRAQM2

(a)

(b)

(c)

(d)

(f)

(g)

(h)

(i)

(j)

(e) (k)










Figure 7

Stations

[H2O]

[Na+]

[W-Na+]

[W-NO3-]

[T-NO3-]

[NOx]

[W-NH4+]

[T-NH4+]

[W-nss-SO42-]

[nss-SO42-]

[SO2](a)

(c)

(e)

(g)

(i)

(b)

(d)

(f)

(h)

(j)

NC FCStations

NC FC




4 NH+

4 and NOminus

3 were within










3 and W-NOminus

3 ) T-NOminus








3











3 ] and [T-NOminus




mixed with NOminus







[H+] + [NH+

4 ] + 2([Ca2+] + [Mg2+

]) + [Na+] + [K+

]


3 ] (3)


3 ] + 2([SO2minus

4 ]

+ [CO2minus

3 ] + [SO2minus

3 ]) + [NOminus

3 ] + [HSOminus

4 ]


4 ][NOminus

3 ] [Clminus] [NH+




Figure 8

Obs

Sim

Obs Obs

(a) (b) (c)


(SLSD=90)





4 Conclusions


4 NH+

4 NOminus






Observed pH 460 518 482 449

Simulated pH 471 536 494 459


469 536 493 457


460 528 487 451

1pH(LSP= 90 )b

0014 0027 00096 00024

1pH(LSP= 99 )c

010 014 0033 00062



4 NH+

4 NOminus


4 NH+

4 and NOminus









Edited by H Tost

References













































Figure 5

ObservationRAQM2

(a)

(b)

(c)

(d)

(f)

(g)

(h)

(i)

(j)

(e) (k)


4 (b) NH+

4 (c) NOminus


4 (h) NH+

4

(i) NOminus




4 andNH+



Figure 6

ObservationRAQM2

(a)

(b)

(c)

(d)

(f)

(g)

(h)

(i)

(j)

(e) (k)










Figure 7

Stations

[H2O]

[Na+]

[W-Na+]

[W-NO3-]

[T-NO3-]

[NOx]

[W-NH4+]

[T-NH4+]

[W-nss-SO42-]

[nss-SO42-]

[SO2](a)

(c)

(e)

(g)

(i)

(b)

(d)

(f)

(h)

(j)

NC FCStations

NC FC




4 NH+

4 and NOminus

3 were within










3 and W-NOminus

3 ) T-NOminus








3











3 ] and [T-NOminus




mixed with NOminus







[H+] + [NH+

4 ] + 2([Ca2+] + [Mg2+

]) + [Na+] + [K+

]


3 ] (3)


3 ] + 2([SO2minus

4 ]

+ [CO2minus

3 ] + [SO2minus

3 ]) + [NOminus

3 ] + [HSOminus

4 ]


4 ][NOminus

3 ] [Clminus] [NH+




Figure 8

Obs

Sim

Obs Obs

(a) (b) (c)


(SLSD=90)





4 Conclusions


4 NH+

4 NOminus






Observed pH 460 518 482 449

Simulated pH 471 536 494 459


469 536 493 457


460 528 487 451

1pH(LSP= 90 )b

0014 0027 00096 00024

1pH(LSP= 99 )c

010 014 0033 00062



4 NH+

4 NOminus


4 NH+

4 and NOminus









Edited by H Tost

References













































Figure 7

Stations

[H2O]

[Na+]

[W-Na+]

[W-NO3-]

[T-NO3-]

[NOx]

[W-NH4+]

[T-NH4+]

[W-nss-SO42-]

[nss-SO42-]

[SO2](a)

(c)

(e)

(g)

(i)

(b)

(d)

(f)

(h)

(j)

NC FCStations

NC FC




4 NH+

4 and NOminus

3 were within










3 and W-NOminus

3 ) T-NOminus








3











3 ] and [T-NOminus




mixed with NOminus







[H+] + [NH+

4 ] + 2([Ca2+] + [Mg2+

]) + [Na+] + [K+

]


3 ] (3)


3 ] + 2([SO2minus

4 ]

+ [CO2minus

3 ] + [SO2minus

3 ]) + [NOminus

3 ] + [HSOminus

4 ]


4 ][NOminus

3 ] [Clminus] [NH+




Figure 8

Obs

Sim

Obs Obs

(a) (b) (c)


(SLSD=90)





4 Conclusions


4 NH+

4 NOminus






Observed pH 460 518 482 449

Simulated pH 471 536 494 459


469 536 493 457


460 528 487 451

1pH(LSP= 90 )b

0014 0027 00096 00024

1pH(LSP= 99 )c

010 014 0033 00062



4 NH+

4 NOminus


4 NH+

4 and NOminus









Edited by H Tost

References
















































3











3 ] and [T-NOminus




mixed with NOminus







[H+] + [NH+

4 ] + 2([Ca2+] + [Mg2+

]) + [Na+] + [K+

]


3 ] (3)


3 ] + 2([SO2minus

4 ]

+ [CO2minus

3 ] + [SO2minus

3 ]) + [NOminus

3 ] + [HSOminus

4 ]


4 ][NOminus

3 ] [Clminus] [NH+




Figure 8

Obs

Sim

Obs Obs

(a) (b) (c)


(SLSD=90)





4 Conclusions


4 NH+

4 NOminus






Observed pH 460 518 482 449

Simulated pH 471 536 494 459


469 536 493 457


460 528 487 451

1pH(LSP= 90 )b

0014 0027 00096 00024

1pH(LSP= 99 )c

010 014 0033 00062



4 NH+

4 NOminus


4 NH+

4 and NOminus









Edited by H Tost

References













































mixed with NOminus







[H+] + [NH+

4 ] + 2([Ca2+] + [Mg2+

]) + [Na+] + [K+

]


3 ] (3)


3 ] + 2([SO2minus

4 ]

+ [CO2minus

3 ] + [SO2minus

3 ]) + [NOminus

3 ] + [HSOminus

4 ]


4 ][NOminus

3 ] [Clminus] [NH+




Figure 8

Obs

Sim

Obs Obs

(a) (b) (c)


(SLSD=90)





4 Conclusions


4 NH+

4 NOminus






Observed pH 460 518 482 449

Simulated pH 471 536 494 459


469 536 493 457


460 528 487 451

1pH(LSP= 90 )b

0014 0027 00096 00024

1pH(LSP= 99 )c

010 014 0033 00062



4 NH+

4 NOminus


4 NH+

4 and NOminus









Edited by H Tost

References















































Observed pH 460 518 482 449

Simulated pH 471 536 494 459


469 536 493 457


460 528 487 451

1pH(LSP= 90 )b

0014 0027 00096 00024

1pH(LSP= 99 )c

010 014 0033 00062



4 NH+

4 NOminus


4 NH+

4 and NOminus









Edited by H Tost

References










































