the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Simulations of 7Be and 10Be with the GEOS-Chem global model v14.0.2 using state-of-the-art production rates
Minjie Zheng
Hongyu Liu
Florian Adolphi
Raimund Muscheler
Zhengyao Lu
Mousong Wu
Abstract. The cosmogenic radionuclides 7Be and 10Be are useful aerosol tracers for atmospheric transport studies. Combining 7Be and 10Be measurements with an atmospheric transport model can not only improve our understanding of the radionuclide transport and deposition processes but also provide an evaluation of the transport process in the model. To simulate these aerosol tracers, it is critical to evaluate the influence of radionuclides production uncertainties on simulations. Here we use the GEOS-Chem chemical transport model driven by the MERRA-2 reanalysis to simulate 7Be and 10Be with different production scenarios: the default production rate in GEOS-Chem based on an empirical approach (denoted as LP67), and two production rates from the CRAC:Be (Cosmic Ray Atmospheric Cascade: Beryllium) model considering only geomagnetic cut-off rigidities for a geocentric axial dipole (denoted as P16) or realistic spatial geomagnetic cut-off rigidity variations due to non-dipole moments of the geomagnetic field (denoted as P16spa). The model results are comprehensively evaluated with a large number of measurements including surface air concentrations and deposition fluxes. The model with the P16spa production can reproduce the absolute values and temporal variability of 7Be and 10Be surface concentrations and deposition fluxes on annual and sub-annual scales, as well as the vertical profiles of air concentrations. Simulations with the LP67 production tend to overestimate the absolute values of 7Be and 10Be concentrations. The P16 simulations suggest less than 10 % differences compared to P16spa but tend to produce a significant positive bias (>20 %) in the 7Be deposition fluxes over East Asia. We find that the deposition fluxes are more sensitive to the production in the troposphere and downward transport from the stratosphere. Independent of the production models, surface air concentrations and deposition fluxes from all simulations show similar seasonal variations, suggesting a dominant meteorological influence. The model can also reasonably simulate the stratosphere-troposphere exchange process of 7Be and 10Be by producing stratospheric contribution and 10Be / 7Be ratio values that agree with measurements. Finally, we illustrate the importance of including the time-varying solar modulation in the production calculation, which can significantly improve the agreement between model results and measurements, especially at mid- and high- latitudes. Reduced uncertainties in the production rates, as demonstrated in this study, improve the utility of 7Be and 10Be as aerosol tracers for evaluating and testing transport and scavenging processes in global models.
- Preprint
(9438 KB) - Metadata XML
-
Supplement
(2823 KB) - BibTeX
- EndNote
Minjie Zheng et al.
Status: final response (author comments only)
-
RC1: 'Anonymous referee comment on gmd-2023-111', Anonymous Referee #1, 12 Jul 2023
This study discusses the use of cosmogenic isotopes 7Be and 10Be as aerosol tracers for studying atmospheric transport. By combining measurements of these isotopes with an atmospheric transport model, it is possible to gain insights into their transport and deposition processes and evaluate the model's performance. The study examines different production scenarios using the GEOS-Chem model driven by the MERRA-2 reanalysis and compares them to a large number of measurements. The results show that simulations considering realistic spatial distributions of geomagnetic cut-off rigidities (P16spa) can accurately reproduce the surface concentrations, deposition fluxes, and vertical profiles of 7Be and 10Be. Simulations with the default production rate (LP67) tend to overestimate concentrations. The study also highlights the importance of including time-varying solar modulation in production calculations, which greatly improves the agreement between model results and measurements, particularly at mid- and high-latitudes. Overall, this research contributes to our understanding of aerosol transport and the role of cosmogenic radionuclides in atmospheric studies.
This work covers a highly relevant and rapidly evolving topic, but I am not sure that it fully aligns with the scope of the journal, since this study involves neither the development of a model nor a specific module of a full model but rather utilizes pre-existing code with only the input data being modified. Additionally, in my opinion, the study requires significant improvement. Currently, the main shortcoming is related to the too-short modeling period (one cannot track the solar cycle over the 6-year period studied here). Furthermore, one of the scenarios (LP67) makes little sense since the model uses the outdated production model and the constant modulation potential for different levels of solar activity. However, there are several modern models available that account for changes in the modulation potential. Introducing a scaling coefficient is not correct in this context, especially considering that the authors are addressing the tiny regional differences in the other two scenarios (P16 and P16spa). Therefore, my main suggestion is to add in the abstract and conclusion a strong recommendation not to use LP67 in future studies (since on the global average, the LP67 production rate is 67% higher compared to those of P16 and P16spa). I also suggest extending the time range for P16 and P16spa modelling to at least 12 years, which is a minimum duration needed to estimate whether the model reproduces the solar cycle. The inclusion of seasonal activity in the study is important, and the authors have addressed this aspect well. It would be beneficial to see results on a longer temporal scale as well. In my opinion, the main emphasis in the article could be placed on the fact that, for the first time, the ability of GEOS-Chem to simulate 10Be has been assessed using measurements considering a proper production model is used. Indeed, this is a significant improvement for the model.
Line 30: It is crucial to highlight in the abstract the percentage of mismatch and provide a recommendation to readers not to use this scenario for future research, as it is unsuitable.
Line 50: “incoming galactic cosmic rays (GCRs)” not only galactic, but also solar cosmic rays too.
Line 54-55: as a reference for stratosphere/troposphere production distribution you can also use Heikkilä et al., 2013 (doi:10.1002/jgrd.50217) and Golubenko et al., 2022 (doi:10.1029/2022JD036726).
Line 56: also, Delaygue et al., 2015 (doi:10.3402/tellusb.v67.28582) discussed how the size of aerosol on beryllium transport.
Line 58: Please, add approximately atmospheric residence times.
Line 76-78: Please, also describe the accuracy of this model and the percentage of disagreement between simulation and real data.
Line 80-81: It is incorrect to state that the model described in this sentence (e.g., CCM SOCOL, EMAC) can only be used as a free-running model. Models of this type can be employed as free-running for periods when reanalysis data is unavailable, they can be used with nudging for periods with reanalysis data. Nudging refers to the utilization of reanalysis data and provides similar capabilities to GEOS-Chem (with the only difference being the choice of reanalyses, such as Merra, ERA, or another dataset). On the other hand, GEOS-Chem is unable to function without reanalysis data, particularly when investigating the transport of cosmogenic isotopes in the past (e.g. T. Spiegl et. al., 2022 doi: 10.1029/2021JD035658). Free-run models, however, can be used in such cases. It's important not to mislead readers. In the articles referenced, the authors employ CCM SOCOL in combination with ERA-Interim.
Line 91: The units of the solar modulation potential is MV, not MeV
Line 93-94: Can you please clarify and rewrite this sentence: "Certain modifications of solar modulation need to be applied in simulations for different years to account for changes in solar modulation".
Line 100-104: It would look more readable if each scenario is formatted as an item. For example:
- “Scenario I: GEOS-Chem default production using an empirical proximation (LP67 production);
- Scenario II: production derived from the “CRAC:Be” model with considering realistic geomagnetic cut-off rigidity;
- Scenario III: a production derived from the “CRAC:Be” model with considering only the dipole-moment of the geomagnetic field and an approximation of the resulting latitudinal variations in the cut-off rigidity (the so-called “Stoermer” cut-off).”
Or maybe it will be clearer if the text description here is changed to an overview of the performed simulations (Table 1 from the supplementary materials).
Line 115: The model description lacks some details about the Beryllium module. As this journal is dedicated to models and their development, it is desired to have more specific information. Currently, there are many references, but there are no concise conclusions.
Line 133-134: please use the same number of decimal places in grid size.
Line 134: 80km à 80 km (space missing)
Line 155: Please clarify is the energy spectrum of cosmic rays Ji is a function of the cutoff energy (Ec) or kinetic energy E?
Line 152: Since the index «i» refers to different types of primary cosmic ray particles would you kindly provide information about the ratio of these particles? Do you consider particles that are heavier than alpha particles? If so, could you please explain how you incorporate them into your analysis?
Line 170-173: Please consider including a reference to the work by Nevalainen, Usoskin et al., 2013 (doi: 10.1016/j.asr.2013.02.020)? This study demonstrates that, on a global scale, there are no significant differences between the two scenarios mentioned – one using the Stoermer equation and the other utilizing the real geomagnetic cut-off rigidity inferred from particle trajectories. However, it does highlight minor discrepancies that may arise on a regional scale.
Line 183-184: Please explain the rationale behind using LP67 if there is no possibility to use the modulation potential for years other than that for 1958. Initially, you focus on such a minor difference between P16 and P16spa suggesting that it may be important (cf. Nevalainen et al., 2013 doi:10.1016/j.asr.2013.02.020). However, a much stronger effect of the solar modulation potential is neglected in the LP67 scenario. This is unacceptable because 1958 was a year of the maximum activity, while the years 2012-2014, analyzed here, were during the period of reduced activity, and the assumption of scaling the solar potential is not physically plausible. Previous studies (already mentioned in the paper) have already demonstrated that LP67 has major inaccuracies, and it is better to use P16 or P16spa. If the GEOS-Chem model does not allow for changing the solar modulation potential, it is a serious drawback that raises doubts about the entire experiment 1. Based on the above, you could perform all three scenarios for the year 1958 (although this may have little sense, as the comparison between LP67 and P16 or P16spa has already been done earlier).
Line 203-209: Again, as in lines 100-104. For readers, it’s not clear when all options look like one sentence. Please use items.
Line 220: Please, add more details about 10Be surface air measurements using in this study.
Line 231-253: I believe that Figure 1 and the discussion are unnecessary here, as already mentioned by the authors in the introduction, the advantage of P16 over LP67 has been demonstrated in previous studies. It does not provide any practical utility in this context since the GEOS-Chem model with input parameters from LP67 cannot be applied to any other time period except for the year 1958. However, the authors here study the period of 2012-2014, when the solar modulation potential was almost half than for 1958.
Line 254-263: I suggest discussing not only the presence of anomalies but also their underlying causes in more detail, based on earlier studies as proposed by Neväläinen et al. (2013), and searching for similar cases.
Figure 2: The colour scale in Figure 2 (especially the upper panel) should be made clearer and possibly add contours to the plot. Additionally, please make the latitude numbering bold, similar to panel (d), to ensure consistency and clarity. The current presentation lacks cohesiveness.
Line 270-271: I would like to emphasize once again that comparing the results of P16 and LP67 is not appropriate because we cannot use the same modulation potential for both 1958 and 2012, as done in the LP67 scenario.
Line 314: The use of scaling coefficients does not make sense because there are currently several models available that account for changes in modulation potential. How did you define these coefficients, and what is its purpose? Once again, I suggest removing this scenario from the study.
Figure 4: Could you please use a consistent color palette throughout the entire work? You chose a palette in Figure 1, and it would be convenient to use the same palette in the rest of the figures. Additionally, the measurements in the figure tend to blend with the background on the overall map. Could they be made slightly brighter (bolder)?
Line 327: The contribution of the stratospheric and tropospheric fractions was described, in particular, by Heikkilä et al. 2009 (doi:10.5194/acp-9-515-2009), so this sentence only confirms her assessment.
Line 452: If you normalize 7 and 10BeLP67 to BeP16spa, please indicate this in the legend of all the figures where a comparison with measurements is presented.
Please make sure to give more attention in the conclusions to the fact that LP67, when compared to P16spa, provides distorted values.
Citation: https://doi.org/10.5194/gmd-2023-111-RC1 - RC2: 'Comment on gmd-2023-111', Tobias Spiegl, 11 Aug 2023
- AC1: 'Comment on gmd-2023-111', Minjie Zheng, 25 Sep 2023
Minjie Zheng et al.
Data sets
data and code for "Simulations of 7Be and 10Be with the GEOS-Chem global model v14.0.2 using state-of-the-art production rates" Minjie Zheng, Hongyu Liu, Florian Adolphi, Raimund Muscheler, Zhengyao Lu, Mousong Wu, and Nønne L. Prisle https://doi.org/10.5281/zenodo.8051729
Minjie Zheng et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 289 | 78 | 11 | 378 | 28 | 5 | 5 |
- HTML: 289
- PDF: 78
- XML: 11
- Total: 378
- Supplement: 28
- BibTeX: 5
- EndNote: 5
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1