We present a new global three-dimensional chemical model for the troposphere, named chemical atmospheric general circulation model (AGCM) for study of atmospheric environment and radiative forcing (CHASER). This model, developed in the framework of the Center for Climate System Research/National Institute for Environment Studies (CCSR/NIES) AGCM, is aimed to study tropospheric photochemistry and its influences on climate. The chemical component of the model simulates the O3-HOx-NOx-CH4-CO photochemical system and oxidation of nonmethane hydrocarbons through 88 chemical and 25 photolytic reactions with 47 chemical species in its present configuration. The model includes emission sources, dry and wet deposition, as well as chemical transformations. Meteorological processes such as transport due to advection, convection, and other subgrid-scale mixing are simulated “on-line” by the dynamical component of the CCSR/NIES AGCM. A detailed evaluation of the model results is presented in a companion paper [Sudo et al., 2002]. An evaluation of the transport scheme adopted in the model suggests that the model is capable of simulating transport associated with convection and boundary layer mixing as well as large-scale advection. The model capability to simulate dry and wet deposition was also evaluated by conducting a simulation of atmospheric lead. The simulated lead distributions are consistent with those observed at the surface, showing the validity of the deposition parameterization adopted in the model.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
 The direct and indirect impact of human activities on the atmospheric environment and climate is one of the biggest concerns in recent atmospheric science. The chemical composition of the atmosphere has been changed by not only increase in anthropogenic emission, but also changes in land use. In addition to well-mixed gases as carbon dioxide (CO2) and methane (CH4), reactive species such as ozone (O3) and its precursors (carbon monoxide CO, nitrogen oxides NOx, nonmethane hydrocarbons NMHCs, etc.) are on the increase [e.g., Crutzen and Zimmermann, 1991]. Ozone, a greenhouse gas as well as CO2 and CH4, is the most important chemical species for tropospheric photochemistry to activate chemical reactions and control the life time of other chemical species (oxidation capacity) through formation of hydroxy radical (OH). The impact of these changes in the atmospheric composition on the atmospheric environment and climate is at the global scale. Therefore, global investigation of the behavior of individual chemical species and each process is needed.
 Global chemical models can easily incorporate and reflect one's suggestions in terms of global effect. Additionally, comparisons between model results and observations show us rightness of the present knowledge about atmospheric chemistry or sometimes suggest other possibilities.
 In this paper, we introduce a new global chemical model of the troposphere named CHASER. This model is based on the Center for Climate System Research (CCSR), University of Tokyo/National Institute for Environmental Studies (NIES) AGCM, which has been developed at the CCSR and the NIES. The CCSR/NIES AGCM has been also used for an on-line global simulation of stratospheric chemistry and dynamics [Takigawa et al., 1999], and for a global simulation of the aerosol distribution and optical thickness of various origins [Takemura et al., 2000]. The principal objective of CHASER is to study the global distributions and budgets of ozone and its precursors. Additionally, CHASER can be used to assess the global impact of changes in the atmospheric composition on climate. CHASER has been already employed in a simulation study of tropospheric ozone changes during the 1997–1998 El Niño event [Sudo and Takahashi, 2001].
 CHASER includes transport, chemistry, deposition, and radiation components. Calculated ozone concentration is used for the radiation and photolysis rate (J-value) calculation. We describe a model overview in section 2. Description and evaluation of the transport component are given in section 3. We introduce the chemistry component and emissions used in this study in section 4 and section 5, respectively. In section 6, the deposition processes in CHASER are described and evaluated. A detailed evaluation of the model results of ozone and related chemical species is presented by Sudo et al. .
2. Model Overview
 The CHASER model is based on the CCSR/NIES AGCM. Basic features of the CCSR/NIES AGCM have been described by Numaguti . The newly implemented physical processes were presented by Numaguti et al. . This AGCM adopts a radiation scheme based on the k-distribution and the two-stream discrete ordinate method [Nakajima and Tanaka, 1986]. A detailed description of the radiation scheme adopted in the AGCM is given by Nakajima et al. . The prognostic Arakawa-Schubert scheme is employed to simulate cumulus convection [cf. Numaguti et al., 1995] (see the description by Numaguti  for further details of the hydrological processes in the model). The level 2 scheme of turbulence closure by Mellor and Yamada  is used for the estimation of the vertical diffusion coefficient. The orographic gravy wave momentum deposition in the AGCM is parameterized following McFarlane . The AGCM generally reproduces the climatology of meteorological fields. In climatological simulations, CHASER uses climatological data of sea surface temperature (SST) as an input to the AGCM. In simulations of a specific time period, analyzed data of wind velocities, temperature, and specific humidity from the European Center for Medium-Range Weather Forecasts (ECMWF) are used as a constraint in addition to SST data of a corresponding year, because it may be difficult to validate just climatological output from the model with observations in a certain period.
 In CHASER, dynamical processes such as tracer transport, vertical diffusion, surface emissions, and deposition are simulated in the flow of the AGCM calculation. The chemistry component of CHASER calculates chemical transformations using variables of the AGCM (e.g., temperature, pressure, humidity). In the radiation component, radiative transfer and photolysis rates are calculated by using the concentrations of chemical species calculated in the chemistry component. The dynamical and physical components of CHASER are evaluated with a time step of 30 min. We have chosen a chemistry time step of 10 min. In this study, we adopted a horizontal spectral resolution of T21 (approximately, 5.6° longitude × 5.6° latitude) with 32 layers in the vertical from the surface up to about 3 hPa (about 40 km) altitude. CHASER uses the σ coordinate system in the vertical. The 32 layers are centered approximately at 995, 980, 950, 900, 830, 745, 657, 576, 501, 436, 380, 331, 288, 250, 218, 190, 165, 144, 125, 109, 95, 82, 72, 62, 54, 47, 40, 34, 27, 19, 11, and 3 hPa, resulting in a vertical resolution of 1 km in the free troposphere and much of the lower stratosphere for an accurate representation of vertical transport such as the stratosphere-troposphere exchange (STE).
 The present version of CHASER calculates the concentrations of 44 chemical species from the surface up to about 20 km altitude. The concentrations of O3, NOx, N2O5, and HNO3 in the stratosphere (above 20 km altitude) are prescribed using monthly averaged output data from a three-dimensional stratospheric chemical model [Takigawa et al., 1999]. For the O3 distribution (>20 km), the data of Takigawa et al.  were scaled by using zonal mean satellite data from the Halogen Occultation Experiment project (HALOE) [Russel et al., 1993; Randel, 1998], since the latest version of the stratospheric chemical model [Takigawa et al., 1999] tends to slightly overestimate the O3 concentrations in the tropical lower stratosphere. Concentrations in the stratosphere (>20km) in the model are reset to those data at each time step.
 Transport is one of the most important processes to simulate the atmospheric photochemistry. Emitted or chemically produced species undergo advection by large-scale wind field and subgrid vertical transport by diffusion and moist convection. In CHASER, advective transport is simulated by a 4th order flux-form advection scheme of the monotonic van Leer [van Leer, 1977], except for the vicinity of the poles. For a simulation of advection around the poles, the flux-form semi-Lagrangian scheme of Lin and Rood  is used. Vertical transport associated with moist convection (updrafts and downdrafts) is simulated by the cumulus convection scheme (the prognostic Arakawa-Schubert scheme). In the boundary layer, equations of vertical diffusion and surface emission and deposition fluxes are solved implicitly.
 It is necessary to validate the model capability for simulations of transport. For this purpose, we have conducted a simple simulation of the distribution of atmospheric radon (222Rn). Radon is emitted from the Earth's surface (mainly from land surface) and decays radioactively with a lifetime of 5.5 days. Surface emission of radon considered here is generally based on the study of Jacob et al. , in which radon emission from land surface is set 1.0 atoms cm−2 s−1 uniformally. Some simulation studies based on this radon emission scenario, however, show an underestimation of the simulated radon concentrations at Mauna Loa by a factor of 2–3 compared to observations, with showing relatively good agreement of simulations with observations at other sites [Jacob et al., 1997; Brasseur et al., 1998]. Although there is a possibility that an insufficient transport in the simulations causes this discrepancy on one side, it can be attributed to a higher emission rate of radon in eastern Asia as suggested by Mahowald et al. . To take this into account, emission rate in eastern Asia (10°S–55°N, 100°E–160°E) is tentatively increased by a factor of 2 in this simulation.
Figure 1 shows the simulated radon distributions for June–July–August (JJA). As can be seen in zonal mean distribution (upper panel), radon is vertically transported from the surface up to the tropopause height associated with convective activities in the Northern Hemisphere. Horizontal distribution of radon in the upper troposphere can be seen in the lower panel of Figure 1. An outstanding high concentration over eastern Asia is due to the doubled emission rate in this region. Transport of radon from northern America and Africa to over the Atlantic is seen. Moreover, long range transport of radon from eastern Asia appears to reach the eastern Pacific region including western America. Figure 2 compares the simulated and the observed radon vertical profiles in western America (California) for June and JJA conditions. The model appears to reproduce the observed radon vertical distribution in the middle-upper troposphere well. The radon maximum seen at 8–10 km altitude is much associated with long range transport from eastern Asia, according to Stockwell et al. . This feature is clearly seen in Figure 3 showing the cross-sectional distribution of calculated radon over 36°N for June. It can be seen that the radon distribution in the middle-upper troposphere is largely affected by transport from eastern Asia through much of the eastern Pacific and western America. In Figure 2, radon concentration is slightly underestimated by the model in 1–3 km altitudes, whereas is overestimated at the surface. This may indicate an insufficient mixing between the planetary boundary layer and the lower troposphere. Figure 4 shows a comparison of calculated and observed seasonal variations of surface radon at several sites. The model appears to reproduce observed radon seasonal cycle well. Both the concentration and the time variability of calculated surface radon are generally high in wintertime when vertical transport of emitted radon is not efficient due to low convective activity. The seasonal cycle of spring-maximum at Mauna Loa is also well reproduced with the doubled radon emission in eastern Asia.
 The chemistry component of CHASER includes 34 tracers (transported) and 16 nontracers (radical species and members of family tracers). Table 1 shows chemical species considered in CHASER. Ozone and nitrogen oxides (NO + NO2 + NO3) are transported as families (Ox and NOx, respectively). The concentrations of nitrogen (N2), oxygen (O2), and water vapor (H2O) are determined from the AGCM calculation. In this study, CH4 is not considered as a tracer because of its long chemical lifetime (8–11 years). In the model, CH4 concentration is assumed to be 1.77 ppmv and 1.68 ppmv in the northern and the southern hemisphere, respectively.
 The present version of CHASER includes 25 photolytic reactions and 88 chemical reactions (Tables 2 and 3). It considers NMHCs oxidation as well as the Ox-HOx-NOx-CH4-CO chemical system. Oxidations of ethane (C2H6), propane (C3H8), ethene (C2H4), propene (C3H6), isoprene (C5H8), and terpenes (C10H16, etc.) are included explicitly. Degradation of other NMHCs is represented by the oxidation of a lumped species named other nonmethane volatile organic compounds (ONMV) as in the IMAGES model [Müller and Brasseur, 1995] and the MOZART model [Brasseur et al., 1998]. We adopted a condensed isoprene oxidation scheme of Pöschl et al.  which is based on the Master Chemical Mechanism (MCM, Version 2.0) [Jenkin et al., 1997]. Terpenes oxidation is largely based on the study of Brasseur et al.  (the MOZART model). Acetone is believed to be an important source of HOx in the upper troposphere and affect the background PAN formation in spite of its low photochemical activity. Acetone chemistry and propane oxidation are, therefore, included in this study, based on the MCM, Version 2.0. Heterogeneous reactions on aerosols may reduce the levels of NOx, HOx, and some RO2 radicals [Dentener and Crutzen, 1993; Jaeglé et al., 1999; Jacob, 2000]. However, heterogeneous reactions on aerosols are not considered in this study. They are being implemented in the next version of CHASER.
 Reaction rates for the reactions listed in Tables 2 and 3 are mainly taken from DeMore et al.  and Atkinson et al.  and Sander et al.  for updated reactions. The quantum yield for O(1D) production in ozone photolysis (J1) is based on the work of Talukdar et al. . The photolysis rates (J-values) are calculated on-line by using temperature and radiation fluxes computed in the radiation component of CHASER. The radiation scheme adopted in CHASER (based on the CCSR/NIES AGCM) considers the absorption and scattering by gases, aerosols and clouds, and the effect of surface albedo. In the CCSR/NIES AGCM, the original wavelength resolution for the radiation calculation is relatively coarse in the ultraviolet and the visible wavelength regions as in general AGCMs. Therefore, the wavelength resolution in these wavelength regions has been improved for the photochemistry in CHASER. In addition, representative absorption cross sections and quantum yields for individual spectral bins are evaluated depending on the optical thickness computed in the radiation component, in a way similar to that of Landgraf and Crutzen . The photolysis rate for the O3 → O(1D) reaction calculated for January and July can be seen in Figure 5.
 CHASER uses an Euler Backward Iterative (EBI) method to solve the chemical reaction system. The method is largely based on the work of Hertel et al.  which increases the efficiency of the iteration process by using analytical solutions for strongly coupled species (e.g., OH-HO2). The chemical equations are solved with a time step of 10 min in this study. Configurations of the chemical scheme such as a choice of species, reactions, and reaction rates are automatically processed by the preprocessor to set up the model through input files. Therefore, the chemical reaction system as listed in Tables 2 and 3 can be easily changed by a user.
 Surface emissions are considered for CO, NOx and NMHCs in the model (Table 4). Anthropogenic emissions associated with industry (e.g., fossil fuel combustion) and car traffic are based on the Emission Database for Global Atmospheric Research (EDGAR) Version 2.0 [Olivier et al., 1996]. NMHCs emissions from ocean are taken from Müller  as in the MOZART model. The geographical distribution of biomass burning is taken from Hao and Liu . The emission rates of NMHCs by biomass burning were scaled to the values adopted in the MOZART model [Brasseur et al., 1998]. The active fire (Hot Spot) data derived from Advanced Very High Resolution Radiometer (AVHRR) and Along Track Scanning Radiometer (ATSR) [Arino et al., 1999] are used as a scaling factor to simulate the seasonal variation of biomass burning emissions. In this study, we estimated the timing of biomass burning emissions, using the hot spot data for 1999 derived from ATSR. We assumed that individual daily hot spots in a model grid cause emissions which decline in a timescale of 20 days in that grid. The temporal resolution for biomass burning emissions is 10 days in this study. Simulated biomass burning emissions in South America have peaks in late August and September (e.g., CO emission, Figure 6). In South Africa, biomass burning emissions begin in May or June near the equator and shift southward with having a peak in October, whereas they begin in July in South America. Consequently, biomass burning emissions in South America are concentrated in August and September in comparison to South Africa. In South America, surface CO concentrations calculated by using this biomass burning emission seasonality have their peaks in September, in good agreement with observations in South America. CO has industrial emission sources as well as biomass burning emission. Figure 7 shows the distribution of CO surface emission. Large CO emission is found in industrial regions (principally America, Europe, China, and India) as emissions of other trace gases. Biomass burning emission is intensive in North Africa (January), in South America, and South Africa (September–October) as also seen in Figure 6. Additionally, there are indirect CO sources from the oxidation of methane and NMHCs (computed in the model). The global CO source from the methane and NMHCs oxidation is estimated at 1574 TgC/yr in CHASER (the detailed budget of the tropospheric CO in CHASER is shown by Sudo et al. ).
Table 4. Global Emissions of Trace Gases Considered in CHASER
Units are TgN/yr for NOx, TgCO/yr for CO, and TgC/yr for NMHCs.
 For NOx, emissions from aircraft and lightning are considered as well as surface emission. Data for aircraft NOx emission (0.55 TgN/yr) are taken from the EDGAR inventory. We assume that lightning NOx production amounts to 5.0 TgN/yr in this study. In CHASER, lightning NOx production is calculated in each time step using the parameterization of Price and Rind  linked to the convection scheme of the AGCM. In the model, the proportions of could-to-ground (CG) flashes and intracloud (IC) flashes are calculated using the cloud top height determined from the AGCM convection scheme, following Price et al.  (NOx production by CG flashes is assumed to be 10 times as efficient as by IC flashes). Computed lightning NOx emission is redistributed vertically by updrafts and downdrafts in the AGCM convection scheme after distributed uniformally in the vertical. As a consequence, computed lightning NOx emission is transported to the upper tropospheric layers and fractionally to the lower layers in the model (leading to C-shape profiles) as studied by Pickering et al. . The distributions of aircraft and lightning NOx emission in the model are shown in Figure 8. The aircraft emission seems to have an importance for the NOx budget in the northern mid high latitudes especially in wintertime. The lightning emission is generally intensive over the continent in the summer-hemisphere. In July, lightning NOx production is most intensive in the monsoon region like southeastern Asia and North Africa where convective activity is high in this season. NOx also has a emission source from soils (5.5 TgN/yr). Soil NOx emission is prescribed using monthly data for soil NOx emission from Yienger and Levy , obtained via the Global Emissions Inventory Activity (GEIA) [Graedel et al., 1993].
 Biogenic emissions from vegetation are considered for NMHCs. The monthly data by Guenther et al. , obtained via the GEIA inventory, are used for isoprene, terpenes, ONMV, and other NMHCs emissions. Isoprene emission and terpenes emission are reduced by 20% to 400 TgC/yr and 102 TgC/yr, respectively following Houweling et al.  and Roelofs and Lelieveld . The diurnal cycle of isoprene emission is simulated using solar incidence at the surface. For terpenes emission, the diurnal cycle is parameterized using surface air temperature in the model [Guenther et al., 1995]. Figure 9 shows the distributions of isoprene emission for January and July in the model. In July, isoprene emission is large through much of the continent in the northern hemisphere, with showing significant values in the eastern United States and eastern Asia.
 Deposition processes significantly affect the distribution and budget of trace gas species (e.g., O3, NOx, HOx). The CHASER model considers dry deposition at the surface and wet scavenging by precipitation.
6.1. Dry Deposition
 In CHASER, dry deposition scheme is largely based on a resistance series parameterization of Wesely  and applied for ozone (Ox), NOx, HNO3, HNO4, PAN, MPAN, ISON, H2O2, CO, CH3COCH3, CH2O, MGLY, MACR, HACET, and peroxides like CH3OOH (see Table 1) in this study. Dry deposition velocities (vd) for the lowermost level of the model are computed as
where ra, rb, rs are the aerodynamic resistance, the surface canopy (quasi-laminar) layer resistance, and the surface resistance, respectively. ra has no species dependency and is calculated using surface wind speed and bulk coefficient computed for the model's lowest level in the AGCM. rb is calculated using friction velocity computed in the AGCM and the Shumid number (calculated with the kinematic viscosity of air and the diffusive coefficient for individual species). Finally, the most important resistance rs is calculated as a function of surface (vegetation) type over land and species using temperature, solar influx, precipitation, snow cover ratio, and the effective Henry's law constant calculated for individual species in the AGCM. rs over sea and ice surface are taken to be the values used by Brasseur et al.  (e.g., vd(O3) = 0.075 cm s−1 over sea and ice). The effect of dry deposition on the concentration of each trace gas in the lowest layer is evaluated together with surface emissions and vertical diffusion by solving the diffusion equations implicitly.
Figure 10 shows the calculated 24-hour average deposition velocities (cm s−1) of ozone in January and July. The values show the deposition velocities calculated for the surface elevation. Deposition velocities of ozone are generally higher than 0.1 cm s−1, except for the high latitudes in winter where solar influx is less intense and much of the surface is covered with snow. In July, ozone deposition velocity ranges from 0.2 to 0.5 cm s−1 over land surface in the northern hemisphere (0.3–0.7 cm s−1 in daytime), in good agreement with the observations [Jacob et al., 1992; Van Pul, 1992; Massman et al., 1994; Ritter et al., 1994]. In the tropical rain forest region (e.g., the Amazon Forest), deposition velocities are high with a range of 0.7–1.2 cm s−1 throughout a year, in agreement with the observations [Fan et al., 1990].
6.2. Wet Deposition
 Wet deposition is considered in two different ways in the model; in-cloud scavenging (rain-out) and below-cloud scavenging (wash-out). A choice of species which are subject to wet deposition is determined from their effective Henry's law constant in standard conditions (Hs, T = 298.15 K). In the present model configuration, in-cloud scavenging is applied for species whose Hs are greater than 102 M atm−1, and additionally below-cloud scavenging is also applied if Hs is greater than 104 M atm−1. In this study, in-cloud scavenging is applied for H2O2, HNO3, HNO4, CH2O, MGLY, HACET, ISON, and peroxides (CH3OOH, C2H5OOH, etc.), with below-cloud scavenging for H2O2, HNO3, and HNO4.
 For in-cloud scavenging, the first-order parameterization of Giorgi and Chameides  is employed. The loss rate β (s−1) due to in-cloud scavenging is calculated by
where P is the precipitation production rate (g cm−3 s−1) due to convective precipitation and large-scale condensation, QL is the liquid water content (g cm−3), is the density of liquid water (=1 g cm−3), R is the gas constant (=0.082 atm M−1 K−1), and H is the effective Henry's law constant (M atm−1) of each species calculated as a function of temperature T (K). P, QL, and T are calculated by the AGCM.
 In the case of below-cloud scavenging for HNO3, HNO4 and H2O2, irreversible scavenging is assumed, and the loss rate associated with below-cloud scavenging is given by
where d is the effective raindrop diameter (cm) calculated according to Mason  and Roelofs and Lelieveld , QR is the raindrop density (g cm−3) determined from the precipitation flux (intensity) calculated for individual model levels in the AGCM, Kg is the mass transfer coefficient of a gaseous molecule to a sphere and calculated by an empirical correlation [e.g., Frössling, 1938] as a function of the raindrop diameter d, the kinematic viscosity of air, the diffusive coefficients, and the terminal velocity of raindrops (computed using an empirical relation to the raindrop diameter d). The below-cloud scavenging scheme is applied with respect to convective precipitation and large-scale precipitation separately.
 The calculated loss rates (day−1) of HNO3 due to wet deposition (in-cloud and below-could scavenging) are shown in Figure 11 (zonal mean) as an example. Wet deposition is efficient in the tropics and the midlatitudes for both seasons, associated with the convective activity and the passage of cyclones (migratory cyclones-anticyclones). In January, high scavenging rates leading to a lifetime of 0.5–1.5 days are simulated over South America, South Africa, and the western Pacific including Indonesia and the north part of Australia in the low latitudes of 0°–20°S (not shown), whereas wet deposition is much less intensive in the air descending region as over the southern Atlantic and the southeastern Pacific because of sparse precipitation. In July, wet deposition is most intensive over southeastern Asia from India to China associated with the monsoon circulation, leading to a lifetime of 0.3–1 days in the middle troposphere.
 For an evaluation of the wet deposition scheme described here, we have conducted a simulation using atmospheric lead (210Pb) as a tracer. This simulation has been performed as an extension of the simulation of 222Rn (section 3), because 210Pb is produced by radioactive decay of 222Rn. 210Pb produced from 222Rn, believed to stick to aerosol surfaces rapidly, was assumed to efficiently removed by wet deposition with the same scavenging lifetime for HNO3 as in many other simulations [Balkanski et al., 1993; Lee and Feichter, 1995; Rehfeld and heimann, 1995; Brasseur et al., 1998]. The dry deposition velocity of 210Pb at the surface is taken to be 0.2 cm s−1 over land surface and 0.05 cm s−1 over sea surface, following Balkanski et al. .
Figure 12 shows a comparison of the mixing ratios of 210Pb calculated and observed at the surface. The seasonal variations of 210Pb are well reproduced by the model for all sites. For Mauna Loa, calculated values are in good agreement with the observation because of our augmentation of radon emission in eastern Asia (see section 3). Although Figure 12 indicates that the model successfully simulates the wet deposition process, it should be noted that there may be uncertainties in the surface emission of radon adopted here and precipitation simulated by the AGCM.
 Wet deposition is one of the most important processes, and is the most difficult process to simulate. Therefore, a further development of the wet deposition scheme is needed. Velders and Granier  show that tropospheric chemistry in atmospheric chemistry transport models has some sensitivities to the wet deposition parameterization used in them. The two kinds of parameterizations for in-cloud and below-cloud scavenging used in this study are basically identical to those adopted in the MOZART model [Brasseur et al., 1998]. We are developing more detailed schemes for wet deposition; for example, below-cloud scavenging linked to the cumulus convection scheme of the AGCM, reemission process of solved species due to reevaporation of raindrops, and deposition on ice particles like cirrus [Lawrence and Crutzen, 1998; Lawrence et al., 1999], though sedimentation (gravitational settling) of ice particles is included in the precipitation production rate P in (2).
 We have presented a new global three-dimensional chemical model for the troposphere, named CHASER. The model has been developed, aimed at studying the global distributions and budgets of tropospheric ozone and related gases, and the radiative effect of tropospheric ozone on climate. The model simulates the major processes involving tropospheric photochemistry such as large-scale and subgrid-scale transport, emissions, deposition, and chemical transformations. Transport, deposition, and other dynamical or physical processes are simulated on-line by the CCSR/NIES AGCM with a typical time step of 30 min. The concentration of chemical species (mainly ozone) calculated by the chemistry component of the model is used for the radiation calculation of the AGCM. In this study, we adopted a horizontal resolution of T21 (5.6° × 5.6°) with a relatively high vertical resolution (32 layers from the surface up to about 3 hPa), though the model can run with a higher horizontal resolution like T42 (2.8° × 2.8°) or higher. Though the T21 horizontal resolution adopted in this study is computationally efficient and may be fairly adequate to simulate global tropospheric photochemistry, we are going to conduct simulations with a higher horizontal resolution for an accurate representation of individual processes (especially of transport) and for detailed analysis and evaluation of the model results.
 In this paper, the transport scheme and the deposition scheme used in the model are evaluated by conducting a simple simulation of 222Rn and 210Pb. The results suggest that the model is capable of simulating long-range transport such as Asian outflow reaching to the United States. Though the evaluation of the deposition scheme shows that the model simulates deposition processes well, more detailed scheme for dry and wet deposition will be implemented in the next version of the model.
 The chemistry component of the model accounts for 88 kinetic reactions and 25 photolytic reactions with 47 chemical species in the present configuration. Heterogeneous reactions on aerosols related to NOx, HOx, and some peroxy radicals (RO2), not considered in the present model version, will be included in the next model version.
 The model results are evaluated and discussed by a companion paper [Sudo et al., 2002].
 We are grateful to M. Takigawa for providing output data from his three-dimensional stratospheric chemical model. We wish to thank K. Toyota and M. Capouet for invaluable comments and discussions. We would like to give special thanks to the late Atusi Numaguti (main developer of the CCSR/NIES AGCM) for many discussions. We dedicate this study to the memory of Atusi Numaguti. This work has been supported by Center for Climate System Research of the University of Tokyo and Frontier Research System for Global Change. Comments on the manuscript by two anonymous reviewers are greatly appreciated.