GCAP 2.0: a global 3-D chemical-transport model framework for past, present, and future climate scenarios

Abstract. This paper describes version 2.0 of the Global Change and Air Pollution (GCAP 2.0) model framework, a one-way offline coupling between version E2.1 of the NASA Goddard Institute for Space Studies (GISS) general circulation model (GCM) and the GEOS-Chem global 3-D chemical-transport model (CTM). Meteorology for driving GEOS-Chem has been archived from the E2.1 contributions to phase 6 of the Coupled Model Intercomparison Project (CMIP6) for the pre-industrial era and the recent past. In addition, meteorology is available for the near future and end of the century for seven future scenarios ranging from extreme mitigation to extreme warming. Emissions and boundary conditions have been prepared for input to GEOS-Chem that are consistent with the CMIP6 experimental design. The model meteorology, emissions, transport, and chemistry are evaluated in the recent past and found to be largely consistent with GEOS-Chem driven by the Modern-Era Retrospective analysis for Research and Applications version 2 (MERRA-2) product and with observational constraints.


. GCAP 2.0 meteorology available at time of publication from the GCAP 2.0 data repository.
or Compute Canada (http://geoschemdata.computecanada.ca/ExtData/). Historical meteorology has been archived for the prein-120 dustrial (1851-1860 C.E.) and recent past (2001-2014 C.E.). In addition, we archive near-future (2040-2049 C.E.) and end-ofthe-century (2090-2099 C.E.) meteorology for seven future scenarios ranging from extreme mitigation to extreme warming (see Sect. 5 for a description of the emission scenarios). In addition, to facilitate comparison of GCAP 2.0 meteorology and composition with observations and traditional GEOS-Chem, we have also performed a recent past simulation in which the E2.1 horizontal winds of the r1i1p1f2 variant were "nudged" to match those of the MERRA-2 reanalysis for 2001-2014 C.E.
125 (Menon et al., 2008). Note that we only nudge the winds and not temperature, humidity or surface pressure as may be done in other models. We urge users to be aware of the challenges involved when interpreting the impact of nudged meteorology on atmospheric composition, especially in the stratosphere (e.g., see Orbe et al., 2020a).

GEOS-Chem
GEOS-Chem (http://www.geos-chem.org) is a global or regional 3-D chemical transport model traditionally driven by as-130 similated meteorology products produced by the NASA Global Modeling and Assimilation Office (GMAO) Goddard Earth Observing System Data Assimilation System (GEOS-DAS). The MERRA-2 science product is generated at 0.5 • latitude by 0.625 • longitude and 72 vertical layers extending from the surface to 0.01 hPa (∼38 layers in the tropical troposphere) and available from 1980 C.E. to the present (Gelaro et al., 2017). There is also a near real-time product (GEOS-FP) available at 0.25 • latitude by 0.3125 • longitude horizontal resolution and available from 2012 C.E., although with periodic changes to the 135 underlying code. Both products are provided at hourly temporal resolution for two-dimensional fields and at 3 h resolution for spatial resolution. Although the underlying meteorology would still be the 2 • latitude by 2.5 • longitude of the GCAP 2.0 meteorology, one can benefit from the finer spatial resolution of the emission inventories.
The second method of running GEOS-Chem is the Message Passing Interface (MPI) parallelized variant utilizing a cubedsphere dynamical core known as GEOS-Chem High-Performance (GCHP; Eastham et al., 2018). GCAP 2.0 meteorology is 170 fully compatible with GCHP, although we refer the reader to Sect. 5.2 about the necessary pre-processing of emissions for GCAP 2.0 runs using GCHP.
Lastly, there exists an adjoint of GCClassic used for inverse modeling and sensitivity applications (Henze et al., 2007). Since the adjoint presently works with MERRA-2 meteorology, the GCAP 2.0 meteorology is also compatible with the adjoint code once the vertical resolution is added, allowing for inverse modeling applications in past and future climates.

4 Meteorology
This section evaluates the GCAP 2.0 meteorology by comparing it to its original CMIP6 simulation, the CMIP6 E2.1 ensemble, and the MERRA-2 reanalysis. Model output contributed to the CMIP6 experiment is archived by an international distributed data repository powered by the Earth System Grid Federation (ESGF) and available online at https://esgf-node.llnl.gov/projects/cmip6.   Second, we note that for researchers interested in studying a "Paris Agreement"-like future world, in which future warming is limited to 2 • C over preindustrial levels, one may use the SSP1-1.9, SSP1-2.6, or SSP4.3.4 scenarios. However, if one wishes to study a future world with the more aggressive goal of limiting warming to 1.5 • C over preindustrial levels, then the only scenario that may be used is SSP1-1.9.
200 Figure 3 shows the temporal evolution of the global annual mean precipitation rate in the simulations. Global precipitation rates are forecast to increase in the coming century due to the temperature-driven increase in surface evaporation and saturation vapor pressures. Unlike surface air temperature, the repeat simulations closely follow the original values and are statistically identical except in the recent past historical simulation, where they are globally higher by 0.9 %. The MERRA-2 reanalysis shows substantially more interannual variability in its global precipitation rates. chemical-transport modeling, which we now summarize. The primary difference between MERRA-2 and E2.1 is the relative importance of stratiform versus convective precipitation. Whereas both models agree on the total precipitation flux to the surface (Fig. S16), MERRA-2 has a higher rate of stratiform condensation (Fig. S33) balanced by a higher rate of stratiform re-evaporation (Fig. S45). In contrast, E2.1 has a higher rate of convective condensation (Fig. S34) balanced by a higher rate of convective re-evaporation (Fig. S44). Furthermore, E2.1 has consistently smaller surface roughness heights over the ocean 215 and vegetated regions than MERRA-2; in contrast, MERRA-2 does not appear to include an orographic component in its surface roughness calculation and therefore has lower surface roughness heights over non-vegetated land surfaces (Fig. S29).
Consequently, the E2.1 simulations have lower planetary boundary layer heights over oceans and heavily vegetated regions (globally ∼200 m lower; Fig. S12) relative to MERRA-2. The E2.1 simulations also have a lower tropopause pressure by approximately 40 hPa (Fig. S23); this has been corrected in version E2.2 of the GCM by moving to a higher vertical resolution 220 (Orbe et al., 2020b). Lastly, E2.1 has a higher fraction of photosynthetically active radiation (PAR) present as diffuse (Fig. S10) rather than direct ( Fig. S11) radiation, which will promote higher levels of biogenic emissions (see Sect. 5.2). Note also that MERRA-2 sets PAR fluxes to zero over water. Therefore, coastal and island cells will underestimate the radiation flux in MERRA-2-driven GEOS-Chem simulations, again with consequences for biogenic emissions. Figure 4 compares the spatial distribution of key surface meteorological variables generated for GCAP 2.0 with their 225 MERRA-2 counterparts. Surface air temperature shows near-perfect agreement in spatial distribution between the E2.1 products and MERRA-2. The E2.1 temperatures are slightly higher than MERRA-2, especially over the Northern Hemisphere's oceans; nudging reduces this difference as previously discussed. Total precipitation in the E2.1 fields has a weaker pattern correlation with MERRA-2 since the free-running model produces a split intertropical convergence zone (ITCZ) in the eastern Pacific (a common issue in free-running GCMs, e.g., see Samanta et al., 2019); nudging the winds corrects the spatial patterns 230 but brings the total precipitation rate out of agreement. The surface zonal wind component shows excellent agreement in their spatial patterns, although the magnitudes are greater in the E2.1 simulations relative to MERRA-2. The free-running GCM underestimates the extent of flow towards the Equator over the eastern ocean basins, which is corrected in the nudging simulation (and may be responsible for the improved ITCZ). The E2.1 simulations also lack the relatively large spatial heterogeneity seen in surface winds over the land ice sheets.  Lower and free tropospheric air temperatures are in agreement between E2.1 and MERRA-2, but the higher tropopause leads to colder temperatures in the upper troposphere with respect to MERRA-2. Nudging the winds removes some of the temperature difference in the extratropical upper troposphere but introduces differences in the free troposphere. Specific humidity agrees between E2.1 and MERRA-2 except for a drier polar free troposphere and northern extratropical surface; nudging leads to an 240 additional drying of the stratosphere. The zonal winds agree well between all simulations, particularly between the MERRA-2 reanalysis and the nudged simulation as to be expected. (2019) are shown as colored lines: SSP1-1.9 (green); SSP1-2.6 (blue); SSP4-3.4 (purple); SSP2-4.5 (brown); SSP4-6.0 (orange); SSP3-7.0 (pink); SSP5-8.5 (red). The shaded blue rectangles indicate periods for which GCAP 2.0 input meteorology is available at the time of publication.

Emissions and Boundary Conditions
This section describes the anthropogenic emission inventories and surface boundary conditions from the CMIP6 experiment that have been processed for use by GEOS-Chem, whether driven by E2.1 or MERRA-2 meteorology (Sect. 5.1). It then 245 evaluates and compares the emission fluxes that are sensitive to meteorology between MERRA-2 and E2.1-driven GEOS-Chem simulations in the recent past (Sect. 5.2). narios with different target radiative forcings for the end of this century. They are hosted on ESGF under the input data sets for Model Intercomparison Projects (input4MIPs) project. They are available from https://esgf-node.llnl.gov/projects/input4mips.

255
These emissions and boundary conditions have been processed for input to GEOS-Chem/GCAP 2.0. They are consistent with those that influenced the climate of the E2.1 simulations used to generate the respective GCAP 2.0 meteorologies.

Historical
Historical anthropogenic emissions in CMIP6 are from the Community Emissions Data System (CEDS, Hoesly et al., 2018).
For surface emissions, we processed version 2017-05-18 of the CEDS inventory available at monthly temporal and 0.5 • spatial 260 resolution for the eight surface sectors listed in Table 2 and the 1850-2014 C.E. period. CEDS is presently the default global surface anthropogenic emissions inventory used by GEOS-Chem, although only species for the full-chemistry simulation have been processed and these emissions are overwritten for many locations by regional inventories. Here, we additionally processed the methane and CO 2 fluxes for those GEOS-Chem specialty simulations. All surface emissions increased exponentially over the historical period except for sulfur dioxide (SO 2 ), whose emissions peaked in the 1980s (Fig. S52). We also processed the three-dimensional CEDS aircraft emissions for input into GEOS-Chem as a CMIP6-compliant alternative to the Aviation Emissions Inventory Code (AEIC, Stettler et al., 2011) source that is the default in GEOS-Chem. We processed version 2017-08-30 of the CEDS aircraft inventory, available for NO, CO, black and organic carbon, SO 2 and ammonia (NH 3 ) at monthly temporal and 0.5 • spatial resolution and 25 vertical levels of equal thickness from the surface to 15 km. We have vertically re-gridded to the native 40-level E2.1 resolution and the reduced stratospheric 47-level MERRA-2 resolution. Global aircraft 270 emissions increased mostly linearly across the historical period beginning ca. 1950 C.E. (Fig. S53).
In addition, we processed the biomass burning emissions from version 1. tercomparison Project (FireMIP) for earlier periods. BB4MIPs provides the total mass flux per species at monthly temporal and 0.25 • spatial resolution. We have re-gridded to 0.5 • spatial resolution and speciated for input to GEOS-Chem using a consistent hydrocarbon speciation scheme with CEDS. Fire emissions from BB4MIPs increase slightly across the historical period. However, interannual variability greatly increases in the second half of the 20 th century (Fig. S54), driving the large interannual variability observed in the total emissions during this period (e.g., Fig. 6b).

280
We have prepared version 1.2.0 of the CMIP6 surface boundary conditions derived from historical observations (Meinshausen et al., 2017) and available at monthly temporal resolution and as 0.5 • latitude bands. These have been re-gridded to 0.5 • global spatial resolution for input into GEOS-Chem. Carbon dioxide, methane, and nitrous oxide have monotonically increased since the preindustrial due to anthropogenic activity, whereas shorter-lived stratospheric ozone-depleting substances peaked around the turn of the last century following their ban under the Montréal Protocol (Fig. 7). 285

Future Scenarios
The future anthropogenic emissions and boundary conditions used by CMIP6 and processed here for GCAP 2.0 are summarized by . In brief, the so-named "Shared Socioeconomic Pathways" (SSPs) are determined from integrated assessment modeling (IAM) of five future societal narratives that may be followed to limit future warming to a target radiative forcing. The nomenclature for characterizing the SSP scenarios is SSPx-y.z (or SSPxyz), where x is the number of the future 290 narrative (1 to 5; Table 3), and y.z represents the target radiative forcing in W m −2 at the end of the 21 st -century ranging from 1.9 (extreme mitigation; low warming) to 8.5 (low mitigation; extreme warming). The names of the five narratives are listed in Table 3, and each employs different assumptions about how global society would achieve the target radiative forcing.
SSP1 assumes low challenges to mitigation and adaptation , SSP2 assumes medium challenges to mitigation and adaptation , SSP3 assumes high challenges to mitigation and adaptation (Fujimori et al.,295 2017), SSP4 assumes low challenges to mitigation and high challenges to adaptation , and SSP5 assumes high challenges to mitigation and low challenges to adaptation .
For GCAP 2.0, we focus on seven scenarios corresponding to Tiers 1 and 2 of the ScenarioMIP experiment: SSP1-1.9, SSP1-2.6, SSP4-3.4, SSP2-4.5, SSP4-6.0, SSP3-7.0, and SSP5-8.5 (O'Neill et al., 2016). Assumptions about future population growth, urbanization, gross domestic production (GDP), energy, land-use, and air pollution trends of the SSP IAM scenarios 300 are described in a series of manuscripts (Crespo Cuaresma, 2017;Dellink et al., 2017;Jiang and O'Neill, 2017;Leimbach et al., 2017;Samir and Lutz, 2017;Bauer et al., 2017;Popp et al., 2017;Rao et al., 2017). Version 1.1 of these scenarios (Gidden et al., 2019) were obtained from input4MIPs on ESGF and available at 0.5 • spatial and monthly resolution for 2015 C.E. and then every 10 years beginning with 2020 C.E. and ending with 2100 C.E. They were processed for input to GEOS-Chem in the same way as the historical emissions and boundary conditions. Linear interpolation was used to develop individual yearly 305 emissions between the available decadal values. In addition to the sectors of Table 2, future CO 2 emissions include a "neg" sector that considers negative emission (i.e., carbon capture).
From the perspective of the simulated climate, the forcing that dominates the end-of-the-century response is the CO 2 abundance. Therefore, CO 2 is the only gas that monotonically increases with future forcing target values (Fig. 7a). The trajectories of the remainder of the well-mixed greenhouse gases, stratospheric ozone-depleting substances, short-lived climate forcers, 310 and air-pollution precursors vary between the SSP scenarios and target forcings (Figs. 6 and 7b-l). For example, methane in and the world population will continue to grow, it is the only emission expected to remain constant or grow into the future.
Otherwise, the narrative assumptions strongly influence the global and regional emission changes. Furthermore, we note that the historical emissions inventories have large amounts of interannual variability (primarily due to biomass burning), whereas 315 the SSP scenarios have very low interannual variability. This highlights the necessity for CTM studies such as those that may be accomplished by GCAP 2.0 that can explore the impact of a wider range of future emission trajectories on air quality, shortlived climate forces, and stratospheric ozone in future warmer climates whose meteorology is primarily driven by changes in CO 2 . It also highlights the necessity for performing simulations long enough to establish robust statistics for chemistry-climate interactions (e.g., Garcia-Menendez et al., 2017). (MEGAN), which responds positively to changes in diffuse photosynthetically active radiation (PAR), recent surface air temperature, and soil root wetness (Guenther et al., 2012). There is an option for emissions to respond to CO 2 abundance as well 335 (Tai et al., 2013). Because E2.1 has a greater proportion of PAR present as diffuse radiation, isoprene emissions in E2.1-driven simulations are about 40 % higher than in the MERRA-2-driven simulation.
Panels d-f of Fig. 8 show a very tight agreement in the spatial pattern and magnitude of the flux of dimethylsulfide (DMS; (CH 3 ) 2 S) produced by marine phytoplankton. Emissions of NMHCs from marine environments are represented as the product of prescribed seawater concentration distributions and sea-to-air transfer velocities calculated via the parameterization of 340 Nightingale et al. (2000a, b). The latter respond to sea-surface temperatures and surface wind velocities.
Panels g-i of Fig. 8 compare the source of mineral dust between the simulations. We use the Dust Entrainment And Deposition (DEAD) scheme for mineral dust evasion (Zender et al., 2003), which responds to changes in surface friction velocity (u * ), roughness height, snow/ice cover and depth, soil wetness, pressure, specific humidity and temperature. Dust mobilization was found in our tests to be extremely sensitive to the meteorology product used, with poor spatial correlation (R ≤ 0.26) and with 345 each meteorology product yielding a different order of magnitude in its global total. Therefore, we have determined respective scaling factors for the E2.1 simulations that bring the present-day global total into agreement with the MERRA-2-driven value.
These are included by default in the GCAP 2.0 run directories for the DEAD dust scheme.
Panels j-l of Fig. 8 show the source of sea salt aerosol. The sea-salt mobilization scheme is described by Jaeglé et al. (2011) and responds to sea-surface temperature, surface wind velocity, and the fraction of sea-ice coverage. There is an excellent  Price and Rind (1992) and global mean lightning flash rates are tuned to climatology in the recent past (Cecil et al., 2014). However, because the spatial and 360 seasonal climatology in the MERRA-2-driven simulations is constrained by satellite observations (Murray et al., 2012), which is not appropriate for a free-running GCM, the spatial patterns differ between the simulations. Therefore, E2.1 overestimates the fraction of lightning in the tropics with respect to the extratropics, and puts too much lightning over South America and Oceania and not enough over Africa. It is also worth emphasizing that we do not know how lightning has changed since the preindustrial or will change in a warming world (e.g., Williams, 2005;Price, 2013;Murray, 2016Murray, , 2018Finney et al., 2018).

365
Panels p-r of Fig. 8 show the source of NO from soil microbial activity. The parameterization is described by Hudman et al. (2012) and responds to surface air temperature, wind speed, soil wetness, cloud fraction, downwelling shortwave radiation, and snow/ice cover. To a lesser degree, lightning can also influence the soil NO source through its impact on nitrate deposition to the soils. The spatial correlation is excellent between the different sources, although the E2.1 magnitude is higher by about 40 %.

370
Lastly, we note the important and variable geologic source of SO 2 . Volcanic emissions of SO 2 in GEOS-Chem are normally prescribed from the Aerosol Comparisons between Observations and Models (AeroCom) point-source inventory (Carn et al., 2015), with data available since 1978 C.E. The CMIP6 experiment did not provide historical or future emission fluxes for volcanism. Instead, input4MIPs provided time series of stratospheric aerosol surface area densities and effective radii with which to force the GCMs. Therefore, when users generate a GCAP 2.0 run directory, they are given the option to select a fixed 375 historical AeroCom year from which to prescribe their volcanic emissions.
This section evaluates the performance of GCAP 2.0 driven by E.21 versus MERRA-2 meteorology for the recent past through comparison with observations. We first evaluate model physics and transport using the "TransportTracers" variant of GEOS-Chem (Sect. 6.1). We then evaluate the standard full chemistry mechanism (Sect. 6.2). resolution (see Fig. 1). Identical initial conditions were re-gridded to each model's respective vertical resolution. The first four years of each simulation were discarded as initialization, with the remaining ten years used for evaluation and statistics. All prescribed emissions were identical between each simulation.

Transport
Model transport and physical processes may be evaluated against observations using the "TransportTracers" variant of GEOS-
Sulfur hexafluoride is a trace gas of anthropogenic origin that is chemically and physically inert on human time scales (lifetime of 3200 yr). It is primarily emitted at the surface in the northern hemisphere (Maiss and Brenninkmeijer, 1998). Its meridional gradient may be used to test the rate of inter-hemispheric mixing (Rigby et al., 2010;Hall et al., 2011) and its 395 vertical gradients may be used to infer the age of air in the stratosphere (Waugh and Hall, 2002;Waugh, 2009). Its meridional gradients may also be used to infer the tropospheric age of air (Waugh et al., 2013 life 1600 yr) found in uranium ores. Its evasion from surface soils is relatively uniform and constant and is as described by Jacob et al. (1997). Its insolubility and time scale of decay make it a useful tracer for diagnosing quick vertical mixing within atmospheric models from boundary layer processes and moist convection (e.g., Allen et al., 1996;Brost and Chatfield, 1989;Considine et al., 2005;Feichter and Crutzen, 1990;Hauglustaine et al., 2004;Jacob and Prather, 1990;Jacob et al., 1997;Lambert et al., 1982;Mahowald et al., 1995;Stockwell et al., 1998).

405
Radiogenic 210 Pb is the chemically-inert decay product of 222 Rn. It is readily taken up by sub-micron aerosol particles and subsequently removed from the atmosphere by deposition or decay (Bondietti et al., 1988;Maenhaut et al., 1979;Preiss et al., 1996;Sanak et al., 1981). Because of its relatively long lifetime (half-life 22.2 yr), nearly all 210 Pb is removed via deposition. As its source from 222 Rn is relatively well known, and there is a global and long-term surface deposition flux inventory (Preiss et al., 1996), it is the standard test for model deposition.

410
Cosmogenic 7 Be is produced by cosmic-ray spallation of N 2 and O 2 , predominantly in the polar upper troposphere and lower stratosphere (Lal et al., 1958). The source of 7 Be is updated for this work to use the parameterization of Usoskin and Kovaltsov (2008). Mean solar activity is assumed (solar modulation potential Φ = 670 MV), leading to an average production rate of 0.065 atoms cm −2 s −1 ; about 60 % in the stratosphere and 40 % in the troposphere. Like 210 Pb, 7 Be is rapidly taken up by sub-micron aerosol particles (Bondietti et al., 1988;Maenhaut et al., 1979;Papastefanou, 2009;Papastefanou and Ioannidou, 415 1996;Sanak et al., 1981). It is subsequently transported until removal by deposition or radioactive decay (half-life 53.3 d). 7 Be has been used to constrain vertical transport, wet deposition fluxes, and stratosphere-troposphere exchange in models (e.g., Allen et al., 2003;Brost et al., 1991;Koch et al., 1996;Liu et al., 2001Liu et al., , 2016Barrett et al., 2012). Table 4 gives the atmospheric budget of the three radionuclides driven by the three meteorological products. Observations were aggregated at model spatial and monthly temporal resolution, compared to that month's SMO value, from which zonal climatologies were determined. Also shown is the value of each simulation sampled and processed as in the obser-

Vertical Mixing -Troposphere
We assess vertical mixing within the troposphere using vertical profiles of 222 Rn and the ratio of 7 Be to 210 Pb in surface air. Figure 10 shows simulated climatological 222 Rn profiles sampled at the month and location of the available observations, 435 also plotted. Observations are scarce and available only at northern extratropical continental locations (Bradley and Pearson, 1970;Nazarov et al., 1970;Wilkening, 1970;Moore et al., 1973;Kritz et al., 1998). In an overly convective atmosphere, the vertical gradient of 222 Rn would disappear. There is a slight overestimate within the boundary layer and underestimate above in all of our simulations, implying a small underestimate in boundary layer ventilation. Our results are comparable to or better than other atmospheric models (e.g., see Fig. 5 of Considine et al., 2005).  Figure 11 shows the annual mean surface mixing ratios of 7 Be, 210 Pb and their ratio in our simulations. Since 7 Be is produced in the upper troposphere/lower stratosphere and 210 Pb is produced near the surface, and because the ratio of 7 Be to 210 Pb is unaffected by deposition, the ratio serves as a useful measure for tropospheric vertical mixing (Koch et al., 1996). A persistent high bias would indicate excessive downward transport and/or insufficient upward transport, assuming no bias in either source. The left column shows the climatological long-term data from the surface monitoring stations of  Figure 12 shows the simulated zonal climatology of the age of air in the stratosphere, which is defined as the mean time since an air mass at a given location was last in the troposphere (Hall and Waugh, 2000). Age of air increases away from the equatorial 455 tropopause where most tropospheric air enters the stratosphere (Holton et al., 1995). We determine age of air by using SF 6 as a chronological tracer (e.g., Waugh and Hall, 2002) to determine the average temporal lag between a mixing ratio at a given location in the stratosphere relative to the tropical tropopause for the period 2005-2014 C.E.
470 Figure 13 shows the annual zonal fraction of 7 Be of stratospheric origin in each simulation. Using E2.1 (MERRA-2) meteorology, we find that 23 % (30 %) of annual average surface 7 Be abundance from 38-51 • N is of stratospheric origin, slightly lower (higher) than the observational constraint of 25 % reported by Dutkiewicz and Husain (1985). This is a dramatic improvement over the earlier GCAP / ICECAP studies in which the stratospheric downwelling source in the E2-driven simulations was greatly overestimated (Murray et al., 2014) and was also seen in the NASA Global Modeling Initiative (GMI) CTM driven 475 by GISS Model II' meteorology (Liu et al., 2016). This possibly reflects improvements in downward mass flux in the GCM associated with the increase in vertical resolution (Fig. 1) and updates to the GEOS-Chem dynamical core code that occurred since the original GCAP/ICECAP.  Be to the few observations that exist (Baskaran et al., 1993;Bleichrodt, 1978;Brown et al., 1989;Dibb, 1989;Du et al., 2008;Harvey and Matthews, 1989;Hasebe et al., 485 1981;Hirose et al., 2004;Igarashi et al., 1998;Narazaki and Fujitaka, 2010;Nijampurkar and Rao, 1993;Olsen et al., 1985;Papastefanou et al., 1995;Schuler et al., 1991;Turekian et al., 1983;Wallbrink and Murray, 1994, and references therein).
The wet deposition flux is increased in the E2.1 simulation with respect to the MERRA-2 simulation, particularly over the mid-latitude oceans, and is the closest simulation to matching the observations.
We find a tropospheric residence time for 210 Pb-containing aerosols against deposition of 8.3 d in MERRA-2 versus 8.8 490 d in the E2.1 simulations (Table 4), an improvement in consistency with respect to GCAP / ICECAP, and within the range of previous estimates of 6.5-12.5 d (Turekian et al., 1977;Lambert et al., 1982;Balkanski et al., 1993;Koch et al., 1996;Guelle et al., 1998b, a;Liu et al., 2001). We find a similar relative increase in the lifetime of 7 Be-containing aerosols of 19.7 d versus 21.7 d, also consistent with earlier findings of 23 d (Koch et al., 1996) and 21 d . Whereas wet deposition in MERRA-2-driven simulations is primarily due to stratiform clouds, it is primarily due to convective clouds in 495  Table 5. Lifetime against oxidation by tropospheric OH (yr).

520
The E2.1 simulations have greater total sources of NO y because of the larger natural sources from lightning and soils (see Sect. 5.2) and a greater flux of NO y transported from the stratosphere from the products of nitrous oxide (N 2 O) oxidation. The NO y speciation between family members is largely consistent between the simulations. The lifetime of NO y in the nudged E2.1 simulation is longer than in the free-running E2.1 simulation, reflecting the reduction in that simulation's global mean precipitation flux.  The simulations have been sampled at the satellite's overpass time and the tropopause was determined online within the model following the thermal definition is used to calculate the partial columns. All simulations well-reproduce the spatial distribution 530 of tropospheric NO 2 (all R ≥ 0.94) and have small global mean low biases, but statistically disagree with the satellite product in the subtropical latitudes. The subtropical disagreement is potentially due to uncertainties in the tropopause height between the various products. Tropospheric columns match over East Asia during this period but are underestimated over North America and Europe.  Lifetime (d) 53.6 ± 0.5 57.9 ± 0.8 59.5 ± 0.8 Annual mean and standard deviation. The percentage of the total is given per source and sink. Table 7 gives the tropospheric emissions and lifetimes for key reduced carbon species in all three simulations. Species primarily emitted from terrestrial plants such as isoprene and monoterpenes have higher emission rates in E2.1 due to the higher diffuse radiation fluxes than in MERRA-2 (see Sect. 4 and Sect. 5.2). Species that are primarily lost via oxidation by OH, such as isoprene, have shorter lifetimes in E2.1 due to the higher OH abundances (see Table 5). In contrast, soluble species such as methanol have longer atmospheric lifetimes in E2.1 due to the lower large-scale stratiform precipitation rates in E2.1 relative 540 to MERRA-2. Terrestrial columns simulated by the E2.1 simulations are higher than the MERRA-2 simulations, reflecting the higher biogenic emissions. All simulations underestimate HCHO columns over the remote ocean aside from the continental outflows of North America and Asia and over the Arctic.

535
550 Table 8 gives the tropospheric budget for carbon monoxide (CO) in all three simulations. The direct emissions of CO from anthropogenic and biomass burning sources between the three simulations are identical by experimental design. However, the chemical source of CO from methane and non-methane hydrocarbon oxidation is higher in the E2.1 simulations due to the higher OH abundances (Table 5). This is balanced by the increased chemical loss of CO by OH. The influence of OH on CO production from short-lived non-methane hydrocarbon species seems to outweigh the influence on CO loss, and consequently 555 there are slightly longer CO tropospheric lifetimes in the E2.1 simulations.  Table 9 gives the tropospheric budget for the odd-oxygen family (O x ≡ O 3 + rapid cycling species) for all three simulations.
The individual budget terms are all are consistent with the range of reported values from the Tropospheric Ozone Assessment Report (TOAR) multi-model assessment (see Fig. 3 of Young et al., 2018), the CMIP6 models that performed interactive tropospheric chemistry (Griffiths et al., 2021), as well as the last extensive tropospheric ozone budget evaluation within the 570 standard GEOS-Chem model (Hu et al., 2017). The E2.1 simulations are on the high end of the previously reported values due to the higher tropopause height in those simulations; the upper troposphere and lower stratosphere regions contribute strongly to each O x budget term due to the rapidly increasing abundances of ozone with altitude there. We point out that the stratosphere-to-troposphere flux calculated using the "residual method" of the other budget terms yields consistent results when we track the mass flux of ozone across the dynamic tropopause in the model. Relative to the earlier GCAP and ICECAP 575 simulations (Murray et al., 2014), the transport of ozone from the stratosphere is dramatically improved.   TCO to OMI/MLS are sensitive to uncertainties in the tropopause location in the satellite product versus the models (Griffiths et al., 2021). The E2.1 TCO are globally higher than the OMI/MLS product by 12 %, reflecting the lower tropopause pressures.
However, the only locations in which it is statistically different regarding interannual variability are over the tropical oceans, where it is biased low. This likely reflects the more vigorous convection in E2.1 that leads to ozone destruction (e.g., Murray et al., 2013). Tropospheric columns in MERRA-2 match the global mean from OMI/MLS but also underestimate the western 600 Pacific and additionally underestimate northern extratropical ozone. All simulations underestimate tropospheric columns in the southern extratropics with respect to OMI/MLS. The models all show excellent agreement with respect to total ozone columns with small positive mean global biases of 4 % and high pattern correlation (all R ≥ 0.96). Total ozone in the tropics is higher in MERRA-2 than E2.1, consistent with differences in the rate of vertical ascent in the tropical pipe implied by the stratospheric age of air comparison (see Sect. 6.1.3). The E2.1 simulation overestimates Antarctic ozone relative to MERRA-2 or the nudged 605 simulation, although not significantly compared to interannual variability. shows the mean bias of the model with respect to the observations. The tropopause was determined in the simulations using the thermal lapse rate for comparison with satellite tropospheric products. The bottom row shows the total column of sulfur dioxide (SO 2 ) in Dobson Units in the simulations versus the second public release of version 3 of the OMSO2e product from OMI on the Aura satellite (Li et al., 2020, doi:10.5067/Aura/OMI/DATA3008).

Particulate Matter
The model has been sampled at the time of the Aura overpasses for comparison to the satellite products. Figures S67-S68 of 615 the supplementary materials provide seasonal detail for AOT and the SO 2 columns. GCAP 2.0 is a one-way offline coupling between the E2.1 version of the NASA GISS GCM frozen for the CMIP6 experiments (Kelley et al., 2020;Miller et al., 2021) and the GEOS-Chem 3-D chemical-transport model (http://www.geos-chem.org; Bey et al., 2001). Additional sub-daily diagnostics were added to E2.1 to archive the same fields as the MERRA-2 reanalysis 630 product (Gelaro et al., 2017) that is normally used to drive GEOS-Chem. We then re-performed one of the atmosphere-only members of the E2.1 contributions to the CMIP6 ensemble, archiving the meteorology necessary for driving GEOS-Chem. The E2.1 meteorology is available at 2 • latitude by 2.5 • longitude and for 40 vertical layers ranging from the surface to 0.1 hPa. At publication time, meteorology is available for the preindustrial (1851-1860 C.E.) and recent past (2001-2014 C.E.), including a recent-past simulation nudged to MERRA-2 to assist users in comparing with observations. Also available is meteorology for 635 the near future (2040-2049 C.E.) and the end-of-the-century (2090-2099 C.E.) for seven future "Shared Socioeconomic Pathway" (SSP) scenarios ranging from extreme mitigation (SSP1-1.9) to extreme warming (SSP5-8.5). In addition, the CMIP6 emissions and surface boundary conditions (Hoesly et al., 2018;van Marle et al., 2017;Meinshausen et al., 2017;Gidden et al., 2019) have been processed for input into GEOS-Chem. GCAP 2.0 is operational in all current variants of the GEOS-Chem model, with all GCClassic run directories and input files provided. All GCAP 2.0 input data is publicly 640 served at http://atmos.earth.rochester.edu/input/gc/ExtData/.
The meteorology was evaluated by comparing to both the original simulation and the MERRA-2 reanalysis for the recent past. Surface air in the repeat simulation is slightly warmer than the original run due to increased calls to the radiation code necessary for archiving shortwave fluxes for input to GEOS-Chem. The E2.1 climatology in the recent past largely agrees with the MERRA-2 climatology for that period with the primary difference being in the relative amount of precipitation in stratiform 645 versus convective clouds as well as a higher tropopause height in E2.1. Emissions that respond to meteorology in GEOS-Chem are slightly higher in the E2.1-driven simulations, including biogenic emissions from terrestrial plants, the lightning and soil microbial sources of reactive nitrogen, and sea-salt evasion. The dust mobilization parameterization was found to be extremely sensitive to resolution and meteorology, and scaling factors have been determined to constrain the global source.
Model physics and transport were evaluated using simulations and observations of sulfur hexaflouride and radionuclides.
In all cases, transport is substantially improved over the original GCAP, and the E2.1-driven simulations perform comparably to the MERRA-2-driven simulations. Most importantly, whereas age of air remains too young in both E2.1 and MERRA-2, the stratosphere-to-troposphere mass flux now yields comparable values, with consistent stratosphere-to-troposphere fluxes of ozone with multi-model means. This is a major improvement over the previous versions of GCAP (Wu et al., 2007;Murray et al., 2014). However, we urge users to be cautious when using these fields for stratospheric chemistry-climate applications.

655
Future simulations will provide CMIP6 meteorology from the 102-layer version of the GISS GCM (E2.2; Rind et al., 2020), which will be better suited for studies of the middle atmosphere; E2.2 includes an interactive quasi-biennial oscillation and improved polar vortex variability including sudden warmings, which could contribute additional dynamical variability that GEOS-Chem may otherwise not see (Orbe et al., 2020b).
Lastly, the chemistry of the model using the CMIP6 emissions and the different meteorologies was evaluated against a suite 660 of satellite products and in situ observations. The E2.1-driven simulations have higher OH and therefore more accurate methyl chloroform and methane lifetimes. Greater biogenic fluxes and higher OH yield greater abundances of oxidation products (e.g., CO) in the E2.1-driven simulations, improving comparison with observations in the southern hemisphere. However, the MERRA-2-driven simulations have a superior representation of free-tropospheric ozone, likely due to the constrained lightning seasonality and distribution as well as the more realistic tropopause pressure in those simulations. All simulations underestimate 665 particulate matter abundances outside of industrialized areas and AOT in most places. Otherwise, model performance was very similar in all simulations.
Code and data availability.