[Comment#1-1] Sun et al. present the development of the new global chemistry–climate coupled model BCC-GEOS-Chem v2.0, along with an evaluation of key atmospheric constituents, including O₃, HCHO, NO2, CO, PM2.5, and OH. The authors also investigate the impacts of aerosol–radiation interactions and aerosol–cloud interactions, two major new features relative to the earlier v1.0 release, on radiative fluxes and cloud properties. Overall, the manuscript is clearly written, well structured, and provides a thorough description of the model updates and performance. I have only a few minor comments that I recommend addressing before publication.

[Response#1-1] Thank you very much for your comments. We have revised accordingly. All of them have been implemented in the revised manuscript. Please see our itemized responses below.

[Comment#1-2] The paper lacks sufficiently detailed explanations for the improvement in BCC-GEOS-Chem v2.0 relative to v1.0 and BCC-AGCM-Chem. For example, line 485 states that the enhanced performance in simulating tropospheric NO2 columns is “likely due to differences in tropospheric chemistry and deposition,” which is too vague. The authors could provide a more explicit mechanistic explanation for the observed improvements. Similar clarifications would strengthen other parts of the model evaluation section.

[Response#1-2] Thank you for pointing it out. We agree that a more explicit mechanistic explanation should be provided for the enhanced performance in simulating atmospheric chemistry. We have added the following analysis in the text:

“Comparison with satellite observations in Figure 6 demonstrate that BCC-GEOS-Chem v2.0 has significant improvements in simulating tropospheric CH₂O, NO₂, and CO. BCC-GEOS-Chem v2.0 shows a negative mean bias to OMI tropospheric CH₂O column of -1.2±1.5×10¹⁵ molecule cm^-2 averaged over 2012-2014, with positive bias at the tropics. In comparison, both BCC-GEOS-Chem v1.0 and BCC-AGCM-Chem show positive bias of over 50% over the Amazon, central Africa, tropical Asia, and the southeastern United States. These two models show 10-50% larger BVOCs emission over the hotspot regions compared to BCC-GEOS-Chem v2.0 despite comparable global total emissions (410 versus 389 Tg yr^-1), thereby partly explaining their high biases. In addition, the updated aromatic chemistry in BCC-GEOS-Chem v2.0 (GEOS-Chem version 13.0.0 onwards) further reduces simulated CH₂O over tropical regions such as central Africa and the Amazon (Bates et al., 2021). BCC-GEOS-Chem v2.0 also shows lower tropospheric CH₂O columns over the mid-latitudes in both hemispheres compared to OMI retrievals, consistent with the other two models. This pattern may be partly attributed to systematic uncertainties in satellite retrievals, as previous studies have shown that CH₂O retrievals are biased low on average by 20–51% relative to aircraft observations across mid-latitude regions (Zhu et al., 2016; Zhu et al., 2020).

For tropospheric NO₂, BCC-GEOS-Chem v2.0 shows no significant global mean bias compared to OMI tropospheric NO₂ column (0.0±1.3×10¹⁵ molecule cm^-2), but this reflects the compensation of negative bias over emission hotspots such as East Asia, India, Western Europe and central Africa, and positive bias over other regions. In comparison, the simulated tropospheric NO₂ column in BCC-GEOS-Chem v1.0 and BCC-AGCM-Chem show substantially high positive biases of 4-7×10¹⁴ molecule cm^-2 (55.4-107.3%) over continental regions, especially over East Asia and India. Given comparable surface NO_x emissions and consistent lightning NO_x parameterization across the three models, the improved performance in simulating tropospheric NO₂ columns is likely driven by differences in NO_x chemistry and nitrate aerosol treatments between the models. Compared to BCC-GEOS-Chem v1.0, BCC-GEOS-Chem v2.0 with up-to-date GEOS-Chem chemical mechanism has incorporated updated aromatic chemistry (GEOS-Chem version 13.0.0 onwards), the photolysis of nitrate (GEOS-Chem version 14.2.0 onwards, not included in this study), as well as the reaction updates in NO_x chemistry. These updates reduce bias in simulated NO₂ column concentrations, as reported in CESM2-GC (Fritz et al., 2022). Besides, the absence of nitrate aerosol chemistry in BCC-AGCM-Chem (Wu et al., 2020) limits a realistic representation of the gas–particle partitioning of reactive nitrogen, thereby leading to the overestimation of tropospheric NO₂ column.

For CO, we evaluate simulated CO concentrations at 700 hPa, where MOPITT satellite has generally high sensitivity (Emmons et al., 2004; Pfister et al., 2005). BCC-GEOS-Chem v2.0 has also substantially reduced the positive bias relative to observed values, averaged 4.9±11.9 ppbv (5.5%) over the globe, compared with the excessive positive CO bias of BCC-GEOS-Chem v1.0 (24.8±12.9 ppbv) and BCC-AGCM-Chem (50.8±14.1 ppbv) with even larger bias over Asia, central Africa, and the East Pacific Ocean. The excessive CO concentrations in BCC-AGCM-Chem are primarily attributable to a severely underestimated atmospheric oxidizing capacity, with the global column-weighted mean OH concentration being nearly 50% lower than that in BCC-GEOS-Chem v2.0 (0.5×10⁶ molecule cm^-3 versus 1.0×10⁶ molecule cm^-3), in addition to higher biomass burning and anthropogenic emissions (1140 Tg yr⁻¹ versus 925 Tg yr⁻¹; Table S1).

Figure 7 evaluates the simulated surface fine particulate matter (PM_2.5) concentrations with the satellite-derived data as introduced in Section 3.2(Shen et al., 2024). The satellite-derived data reveal high PM_2.5 concentration over East Asia and India due to intensive anthropogenic emissions, and over northern and central Africa due to mineral dust and biomass burning emissions. BCC-GEOS-Chem v2.0 reproduces the overall spatial distributions and magnitude of observed PM_2.5 concentrations, though it tends to marginally underestimate concentrations in the hotspot regions (e.g., Amazon, central Africa, and India), which is most likely due to the coarse model resolution, and the uncertainties of biomass burning (Reddington et al., 2017). BCC-GEOS-Chem v2.0 demonstrates superior performance in simulating PM_2.5 concentration compared with BCC-GEOS-Chem v1.0, which exhibits pronounced overestimation in East Asia, India, the Arabian Peninsula, and Northern Africa, and BCC-AGCM-Chem, which significantly underestimates PM_2.5 concentrations in major emission hotspot regions. This improvement is likely due to its better performance in simulating the gas-phase precursor as shown above (Figure 6), as well as its more comprehensive representation of the aerosol chemical mechanism. The large negative bias in BCC-AGCM-Chem is mainly due to the absence of nitrate aerosol chemistry and the incomplete representation of secondary organic aerosol formation (Wu et al., 2020).”

Additional bibliographical references:

Zhu, L., Jacob, D. J., Kim, P. S., Fisher, J. A., Yu, K., Travis, K. R., Mickley, L. J., Yantosca, R. M., Sulprizio, M. P., De Smedt, I., González Abad, G., Chance, K., Li, C., Ferrare, R., Fried, A., Hair, J. W., Hanisco, T. F., Richter, D., Jo Scarino, A., Walega, J., Weibring, P., and Wolfe, G. M.: Observing atmospheric formaldehyde (HCHO) from space: validation and intercomparison of six retrievals from four satellites (OMI, GOME2A, GOME2B,OMPS) with SEAC4 RS aircraft observations over the southeast US, Atmos. Chem. Phys., 16, 13477–13490, https://doi.org/10.5194/acp-16-13477-2016, 2016.

Zhu, L., González Abad, G., Nowlan, C. R., Chan Miller, C., Chance, K., Apel, E. C., DiGangi, J. P., Fried, A., Hanisco, T. F., Hornbrook, R. S., Hu, L., Kaiser, J., Keutsch, F. N., Permar, W., St. Clair, J. M., and Wolfe, G. M.: Validation of satellite formaldehyde (HCHO) retrievals using observations from 12 aircraft campaigns, Atmospheric Chemistry and Physics, 20, 12329–12345, https://doi.org/10.5194/acp-20-12329-2020, 2020.

[Comment#1-3] In line 290, methane is prescribed, meaning that CH4-OH chemical feedbacks are not represented. The authors could discuss the implications of this limitation and comment on whether incorporating CH₄ interactive chemistry is planned for future model development.

[Response#1-3] Thank you very much for the comment. We apologize for the unclear description of methane in the original manuscript. BCC-GEOS-Chem v2.0 prescribes methane concentrations (as well as long-lived greenhouse gases) only at the surface as a boundary condition, but allows its chemistry and transport within the model. Surface methane is prescribed primarily because its global emissions and sinks remain highly uncertain and strongly imbalanced, as documented in recent studies (e.g., Lu et al., 2021; Saunois et al., 2025). Incorporating interactive surface methane emissions would require a very long spin-up period to reach a steady state, which would substantially increase the computational cost and complicate the interpretation of tropospheric chemistry performance. Prescribing surface methane allows us to focus on the evaluation of short-lived climate forcers (SLCFs) under well-constrained boundary conditions, while avoiding additional uncertainties associated with methane source–sink imbalances. We acknowledge that this configuration may partly hinder the CH₄–OH feedback, and we have clarified this limitation and its implications in the revised manuscript.

“Among the gaseous absorbers, BCC-GEOS-Chem v2.0 prescribes surface mixing ratios of long-lived greenhouse gases (e.g. N₂O, CFC11, and CFC12) and methane using CMIP6 historical forcing data (Section 3), but allows the chemistry and transport of these gases. The use of prescribed surface methane concentrations instead of methane emissions is to ensure a realistic methane distribution in the model, as there are substantial uncertainties associated with its poorly constrained global methane source–sink balance (Lu et al., 2021; Saunois et al., 2025). However, such configuration limits the representation of full methane-OH feedback in the model. H₂O and ozone are not constrained and are fully diagnosed in BCC-GEOS-Chem v2.0.” (Section 2.4)

We have also added a statement in the Conclusion section:

“The feedbacks through methane-OH interactions are not fully represented in the current model configuration and should be addressed by introducing unbiased methane sources.”

Additional bibliographical references:

Lu, X., Jacob, D. J., Zhang, Y., Maasakkers, J. D., Sulprizio, M. P., Shen, L., Qu, Z., Scarpelli, T. R., Nesser, H., Yantosca, R. M., Sheng, J., Andrews, A., Parker, R. J., Boesch, H., Bloom, A. A., and Ma, S.: Global methane budget and trend, 2010–2017: complementarity of inverse analyses using in situ (GLOBALVIEWplus CH4 ObsPack) and satellite (GOSAT) observations, Atmospheric Chemistry and Physics, 21, 4637–4657, https://doi.org/10.5194/acp-21-4637-2021, 2021.

Saunois, M., Martinez, A., Poulter, B., Zhang, Z., Raymond, P., Regnier, P., Canadell, J. G., Jackson, R. B., Patra, P. K., Bousquet, P., Ciais, P., Dlugokencky, E. J., Lan, X., Allen, G. H., Bastviken, D., Beerling, D. J., Belikov, D. A., Blake, D. R., Castaldi, S., Crippa, M., Deemer, B. R., Dennison, F., Etiope, G., Gedney, N., Höglund-Isaksson, L., Holgerson, M. A., Hopcroft, P. O., Hugelius, G., Ito, A., Jain, A. K., Janardanan, R., Johnson, M. S., Kleinen, T., Krummel, P., Lauerwald, R., Li, T., Liu, X., McDonald, K. C., Melton, J. R., Mühle, J., Müller, J., Murguia-Flores, F., Niwa, Y., Noce, S., Pan, S., Parker, R. J., Peng, C., Ramonet, M., Riley, W. J., Rocher-Ros, G., Rosentreter, J. A., Sasakawa, M., Segers, A., Smith, S. J., Stanley, E. H., Thanwerdas, J., Tian, H., Tsuruta, A., Tubiello, F. N., Weber, T. S., Van Der Werf, G., Worthy, D. E., Xi, Y., Yoshida, Y., Zhang, W., Zheng, B., Zhu, Q., Zhu, Q., and Zhuang, Q.: Global Methane Budget 2000–2020, https://doi.org/10.5194/essd-2024-115, 6 June 2024.

[Comment#1-4] Lines 169–176: The manuscript states that the MOM ocean component is coupled into the system, but its spatial resolution differs from the atmospheric component. The authors could explain how this coupling is handled and how the resolution mismatch is treated.

[Response#1-4] Thank you very much for the comment. The atmosphere, land, ocean, and sea ice components interact through the National Centre for Atmospheric Research (NCAR) flux coupler. Surface fluxes computed on one component’s native grid (such as momentum, heat, and freshwater fluxes from the atmosphere, or sea-surface temperature and sea-ice properties from the ocean) are conservatively interpolated onto the other component’s grid. This coupling is achieved using area-weighted or bilinear interpolation schemes that preserve the global integrals of energy, momentum, and mass. We have added the following statements in Section 2.1:

“The above components interact through bidirectional flux exchanges of momentum, energy, water, and carbon facilitated by the coupler, which is adapted from National Centre for Atmospheric Research (NCAR) flux coupler (Craig et al., 2012; Wu et al., 2019). The coupling under different resolution is achieved using area-weighted or bilinear interpolation schemes that preserve the global integrals of energy, momentum, and mass.”

Additional bibliographical references:

Craig, A. P., Vertenstein, M., and Jacob, R.: A new flexible coupler for earth system modeling developed for CCSM4 and CESM1, The International Journal of High Performance Computing Applications, 26, 31–42, https://doi.org/10.1177/1094342011428141, 2012.

[Comment#1-5] Section 2: The authors could clearly specify which components are actively coupled in BCC-GEOS-Chem v2.0. For example, although the interactive land component is included, dry deposition does not use interactive land types or LAI. Clarifying exactly which processes are interactively coupled would avoid confusion.

[Response#1-5] Thank you for pointing it out. We have added statements as presented in [Response #1-4] to clarify that the atmosphere, land, ocean, and sea ice components are actively coupled in BCC-GEOS-Chem v2.0. Specifically for dry deposition, we offer two options, one of which enables direct utilization of land types and LAI consistent with the interactive land module AVIM. We have modified the following statement in Section 2.3.3:

“The scheme requires input of geography data such as land types and leaf area index (LAI) to yield surface roughness and canopy resistance. Here we provide two options in BCC-GEOS-Chem v2.0. Firstly, we can obtain archived land types (Olson et al., 2001) and the LAI data for each land type from Moderate Resolution Imaging Spectroradiometer (MODIS) used in GEOS-Chem v14.0.1. Alternatively, we can obtain the land types and LAI information from the BCC-AVIM through the coupler, and then computes dry deposition through the GEOS-Chem routines, as done in BCC-GEOS-Chem v1.0. The latter option requires reconciliation of land types (73 categories) used in GEOS-Chem with those used in BCC-AVIM (22 categories) (Lu et al., 2020). The key advantage of the latter option is that the land surface parameters used for dry deposition align with those generated by the land surface model, thereby improving its utility for studying atmosphere-land surface interactions. However, uncertainties inherent to the land surface model simulations may propagate into atmospheric chemistry calculations. Both options can be applied in BCC-GEOS-Chem v2.0. In this study, we employ the first option as it offers more detailed vegetation classifications (73 vs 22 categories), ensuring improved compatibility and representation within the GEOS-Chem dry deposition parameterization.” (Section 2.3.3)

[Comment#1-6] Section 3.1: Please provide the time steps used for chemistry, physics, and other components in the simulations.

[Response#1-6] We have added the following table in the supplement.

              Table S4 Time steps for each component in T42L26/T159L72 simulations.

[Comment#1-7] Line 450 attributes the high O3 bias in BCC-GEOS-Chem v2.0 to the updated aromatic chemistry introduced in GEOS-Chem v13.0.0 and later. However, Figure 5 shows that the model performs better for several other pollutants. The authors could discuss the overall impact of this updated chemical mechanism; does it improve overall model skill, degrade it, or introduce compensating effects across different species?

[Response#1-7] Thank you for pointing it out. We have added the following statements to discuss the overall impact of this updated chemical mechanism and further explore the potential causes of the high ozone biases from multiple perspectives, including chemical mechanisms, deposition processes, and representative issues when comparing model results with site measurements in Section 4.1. We have also improved the discussions on the model performance of other pollutants, as shown in [Response#1-2].

“At the surface, however, all three models tend to overestimate the surface ozone, which has been a long-standing issue among global chemical models (Gao et al., 2025). Figure 5 shows the scatter plots of the simulated and observed surface ozone concentrations over Europe, Asia, and the US over 2012-2014. We only use remote or rural sites here as a global model at a coarse resolution of ~2.8° is difficult to resolve pollutant levels at urban sites. BCC-GEOS-Chem v2.0 tends to overestimate surface ozone levels, especially in Asia with high biases of 10-20 ppbv, these high biases are not prominent in BCC-GEOS-Chem v1.0 and other previous studies using earlier GEOS-Chem model version (e.g. version 11).

This discrepancy reflects complicated joint effects of the updated chemical mechanism. For instance, the integration of updated aromatic chemistry in GEOS-Chem models from version 13.0.0 onwards elevate surface ozone concentrations by at least 5 ppbv in eastern China (Bates et al., 2021; Lu et al., 2025). Moreover, the removal of sea salt aerosol debromination in GEOS-Chem models from version 13.4.0 onwards, leading to enhancement in ozone concentrations by up to 10 ppb especially over the Southern Ocean. Other significant updates in heterogeneous chemistry (e.g. updates in N₂O₅, NO₃, NO₂ reactive uptake, nitrate photolysis) can further modulate ozone concentrations through their impacts on ozone precursors. The larger ozone positive bias in BCC-GEOS-Chem v2.0 relative to BCC-GEOS-Chem v1.0 may also attribute to lower dry deposition velocity in the model, as will be discussed in Section 4.3. The representative issue when comparing model results with site measurements are also emerges. Even though the comparisons are limited to rural or remote sites, the low resolution of these models (~200km) cannot resolve the heterogeneity of ozone precursors, thus lead to artificial mixing and biased ozone production efficiency (Wile and Prather, 2006; Yu et al., 2016; Young et al, 2018). In addition, these models are difficult to represent local meteorological conditions particularly over complex terrain at such coarse resolution, further limit their ability to capture site-level ozone concentrations. We will demonstrate in Section 6 that increasing the spatial resolution can significantly reduce the bias in simulated surface pollutant concentrations at site level.”

Thank you very much for your comments. We have revised accordingly. All of them have been implemented in the revised manuscript. Please see our itemized responses in the supplement.