the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 01 Oct 2025
Application and Evaluation of CRACMM V1.0 Mechanism in PM2.5 Simulation Over China
Abstract. Chemical mechanisms are one of the major sources of bias in chemical transport model simulations, making their improvement a critical step towards enhancing model performance and supporting air quality management and research. In this study, a newly developed chemical mechanism, the Community Regional Atmospheric Chemistry Multiphase Mechanism (CRACMM), integrated into the Community Multiscale Air Quality (CMAQ) modeling system, was evaluated through comparison with two traditional chemical mechanisms, CB6r3_ae7 and Saprc07tic_ae7i, for China. Sensitivity simulations related to precursor reactive organic carbon (ROC) emissions were conducted to investigate the key driving factors of PM2.5 formation. The results show slight differences in the correlation coefficient (R), mean bias (MB), and normalized mean bias (NMB) values for the three chemical mechanisms when using the traditional primary organic aerosol (POA) inventory. However, when using the full volatility emission inventory, CRACMM shows improvements in R, MB, and NMB values in some regions. CRACMM predicts higher PM2.5 concentrations during spring, summer and autumn, mainly due to enhanced secondary organic aerosol (SOA) formation driven by increased precursor emissions. Benzene–toluene–xylene (BTX) species and semi-volatile organic compound (SVOC) emissions significantly contributed to PM2.5 formation in CRACMM. The SOA from BTX emissions accounts for nearly 50 % of the PM2.5 changes, while intermediate-volatility organic compounds (IVOC) and SVOCs emissions mainly affect PM2.5 concentrations through SOA formation. These results indicate that CRACMM, when using the full volatile inventory, can effectively compensate for the underestimation of PM2.5 mass that may occur with traditional POA treatment, particularly in regions with high photochemical activity and abundant S/IVOC precursors.
- Preprint
(2256 KB) - Metadata XML
-
Supplement
(2085 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2025-3627', Anonymous Referee #1, 09 Dec 2025
-
RC2: 'Comment on egusphere-2025-3627', Anonymous Referee #2, 15 Dec 2025
In this manuscript, the authors use the CMAQ model to evaluate simulations of PM2.5 using different chemical mechanisms. The work provides a detailed description of the CRACMM mechanism and its implementation, and the authors clearly invested substantial effort in this work. The topic is timely and relevant to the air quality modeling community. I recommend publication of this manuscript once the following comments are addressed.
Major Comments
- Throughout the manuscript, the authors discuss the performance of the CRACMM mechanism using various statistical metrics. However, in many cases the discussion lacks reference to relevant previous work that would help contextualize these results. In addition, statements such as “performance is good in Region 1” or “performance is moderate in Region 2” are presented without further interpretation. I encourage the authors to not only report these findings but also provide explanations for the such differences in performance, ideally supported by literature and known chemical mechanisms and meteorological factors, using additional figures/tables.
- It is my opinion that the introduction section consists of somewhat detailed description of the different chemical mechanisms. I would encourage the authors to move some of the text to a description section or to the supplement and focus more on the motivation of this work. Possible improvements could include: (a) discussing existing literature on air quality modeling over China (b) the region-specific limitations of using a (relatively) simpler (or older) chemical mechanism (c) any computational cost related arguments (d) possible usage of complicated chemistry schemes in other regional scale models (e) potential to use an advanced chemical mechanism in global models (with an advancement in high performance computing).
- I agree with the ‘major comment’ of Reviewer 1, and would recommend the authors to add a paragraph (and a figure/table) carefully explaining their choice.
Additional Comments
- L22: “…two traditional chemical mechanisms, CB6r3_ae7 and Saprc07tic_ae7i, for China.” Kindly consider providing a longer name.
- L25: “…show slight differences in the correlation coefficient…” Please provide a number.
- L28: Same comment as above.
- L41: Please consider citing a couple of more recent papers.
- L56: ‘CMAQ’ appears for the first time in the main text. Full name and citation needed.
- L114: Kindly add one of two citations to which the reader may refer to better understand the importance of these regions.
- L118: Kindly mention the physics and chemistry timesteps. I would also recommend adding a table in the supplement mentioning the different parameterizations used in the study.
- L128: Please consider removing the link and adding a citation for WRF.
- L135: Kindly mention the nominal resolution of the emissions and the sectors included.
- L139: Consider removing the link or move it to the data availability section.
- Table S1: Kindly mention about E/L in the caption.
- Table S2: Consider adding the numbers for the different regions in China.
- All the lat-lon figures seem crowded. Kindly consider removing some of the background colors.
- L226: I recommend adding a couple of maps showing monthly average precipitation and monthly average PBL (or histograms if the authors prefer that).
- Figure 2 (and other similar Figures): The choice of color scale is such that extreme low PM2.5 (0-5 µg/m3) is coinciding with the white background and making it difficult to interpret the figures. Please fix the colorscale.
- L236-L248: The current description could be improved. The correlation co-efficient is one statistical metric. I would encourage to authors to shorten this description and focus on the scientific understanding. Additionally consider adding a couple of sentences in the supplement describing how dust concentrations are calculated in the model (an empirical function of wind speed?)
- In the supplement, please add an equation for R and NMB. I also encourage the authors to report Normalized Root Mean Squared Error.
- L286-287: I recommend adding evidence for this conclusion.
- L299: Fix the sentence.
- Table 2: It is not clear how the recommended benchmark is calculated. It seems that the correlation is lower than the ‘recommendation’. I would encourage the authors to add additional relevant literature and discussion.
- L331: Evidence needed.
- Some of the 4 panel figures have different color scales for different subplots. Please fix the colorscale or mention about different ranges in the caption.
- Section 3.3.2: A brief description of the C* and it’s importance could be beneficial for the readers.
- L462: Chang et al. – Fix.
- L443: A longer description of the species’ is recommended.
- I would recommend the authors to add a simple flowchart to demonstrate different species in different mechanisms and their sources and sinks.
Citation: https://doi.org/10.5194/egusphere-2025-3627-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 225 | 78 | 23 | 326 | 50 | 15 | 18 |
- HTML: 225
- PDF: 78
- XML: 23
- Total: 326
- Supplement: 50
- BibTeX: 15
- EndNote: 18
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
General comment
This manuscript presents an evaluation of the model CMAQ for three chemical mechanisms with respect to PM2.5 concentration. It complements a comparable evaluation for the model performance in simulating ozone. Model predictions are sensititive to the chemical scheme used. Additionally, the benefit of accounting for evaporation (and reactions) of semivolatile compounds from POA is shown. The manuscript is well written and deserves publication. I recommend that the issue I raise in my main comment below is addressed satisfactorily.
Major comments
- The authors rightly acknowledge the limitations and uncertainties stemming from emissions-to-species mapping. However, for the limitations related to mapping L/S/IVOC emissions the authors refers only to Chang et al. (2022) without elaborating how their own choice of the 15 CRACMM species and Emission strengths might affect the results. A discussion on this aspect is needed in my opinion.
Minor comments
- Sect. 2.2.2
It is not clear how the mapping of LVOC, SVOC and IVOC to CRACMM is done. Also in the Supplement no information is given regarding how the Emission strengths of species are assigned. Both Woody et al. (2016) and Chuang et al. (2022) provide emissions by source category, at least in the main text. This lack of information affects the reproducibility of the results.
- in Table 1 "Full volatile inventory" but somewhere else "Full volatility inventory". Is it volatile or volatility?
- l 415-416
I do agree that "NO2 can fully dissolve into water". NO2 has very low solubility. Maybe the authors intended HNO3 considering the following sentence on ammonium salts.
- the hyperlink for https://doi.org/10.5281/zenodo.16791307 is wrong: it contains the string .Suetal