On the simulations of aerosol pH in China using WRF-Chem (v4.0): sensitivities of aerosol pH and its temporal variations in haze episodes

Ruan, Xueyin; Zhao, Chun; Zaveri, Rahul A.; He, Pengzhen; Wang, Xinming; Shao, Jingyuan; Geng, Lei

doi:https://doi.org/10.5194/gmd-2022-62

Preprints

https://doi.org/10.5194/gmd-2022-62

Preprints

Submitted as: model evaluation paper

21 Mar 2022

Submitted as: model evaluation paper | 21 Mar 2022

Status: this preprint is currently under review for the journal GMD.

On the simulations of aerosol pH in China using WRF-Chem (v4.0): sensitivities of aerosol pH and its temporal variations in haze episodes

Xueyin Ruan¹, Chun Zhao^1,2,3, Rahul A. Zaveri⁴, Pengzhen He⁵, Xinming Wang⁶, Jingyuan Shao⁷, and Lei Geng^1,2,3 Xueyin Ruan et al.

Show author details

¹School of Earth and Space Sciences, University of Science and Technology of China, Hefei 230026, Anhui, China
²CAS Center for Excellence in Comparative Planetology, University of Science and Technology of China, Hefei 230026, Anhui, China
³Frontiers Science Center for Planetary Exploration and Emerging Technologies, University of Science and Technology of China, Hefei, China
⁴Atmospheric Sciences and Global Change Division, Pacific Northwest National Laboratory, Richland, WA 99352, USA
⁵School of Environment and Tourism, West Anhui University, Lu'an 237012, Anhui, China
⁶Guangzhou Institute of Geochemistry, Chinese Academy of Sciences, Guangzhou 510640, Guangdong, China
⁷Flight branch, Civil Aviation University of China, Tianjin 300300, China

Received: 04 Mar 2022 – Discussion started: 21 Mar 2022

Abstract. Precise estimation of aerosol pH in chemical transport models (CTMs) is critical to aerosol modeling and thus influencing policy development. We reported WRF-Chem simulated PM_2.5 pH over China during a period with heavy haze episodes in Beijing, and explored the sensitivity of the modeled aerosol pH to factors including emissions of nonvolatile cations (NVCs) and NH₃, aerosol phase state assumption and heterogeneous production of sulfate. We found default WRF-Chem could predict spatial patterns of PM_2.5 pH over China similar to other CTMs, but with generally lower pH values largely due to the underestimates of alkaline species (NVCs and NH₃) and the difference in thermodynamic treatments between different models. Increasing NH₃ emissions in the model would improve the modeled pH in comparison with offline thermodynamic model calculations of pH constrained by observations. In addition, we found that aerosol phase state assumption and heterogeneous sulfate production are important in aerosol pH predictions for regions with low relative humidity (RH) and high anthropogenic SO₂ emissions, respectively. These factors should be better constrained in model simulations of aerosol pH in the future. Analysis of the modeled temporal trend of PM_2.5 pH in Beijing over a haze episode revealed a clear decrease in pH from 5.21 ± 0.88 in clean period to 3.56 ± 0.49 in heavily polluted period. The increased acidity in more polluted conditions is largely due to the formation and accumulation of secondary species including sulfuric acid and nitric acid, even though being modified by alkaline species (NVCs, NH₃). Our result suggests that NO₂ oxidation is unlikely to be important for heterogeneous sulfate production in Beijing haze as the effective pH for NO₂ oxidation of S(IV) is at higher pH of ~6.

Download & links

Preprint (PDF, 1937 KB)

Supplement (2231 KB)

Xueyin Ruan et al.

Status: final response (author comments only)

RC1: 'Comment on gmd-2022-62', Shaojie Song, 25 Apr 2022

Geoscientific Model Development Discussions: “On the simulations of aerosol pH in China using WRF-Chem (v4.0): sensitivities of aerosol pH and its temporal variations in haze episodes”. Ruan et al. investigates the aerosol pH simulations in WRF-Chem model focusing on haze episodes in Beijing. Overall, the paper is scientifically sound, clearly written and easy-to-follow. I only have a few minor comments:

Line204: WRF-Chem

Section4.1.1: I think the minimal response to elevated NVCs in the nearly neutral cases may be related to the role of carbonate. At least in E-AIM and ISORROPIA, the aerosol pH cannot get very high (8 or 10) because of the buffer role of carbonate.

Section4.3: It is interesting to see a systematic difference between ISORROPIA and MOSAIC even when the model inputs are the same. The authors list several possible contributing factors including AWC, phase-partitioning method, activity coefficients and solution method. I suggest the authors to investigate this issue further: (1) is there any difference in the predicted gas-phase ammonia concentrations between the two models? This can give some clues on the contribution of phase partitioning method; (2) the authors note the large difference in AWC in the paper. It seems that both ISORROPIA and MOSAIC use the ZSR method. ISORROPIA uses a look-up table for hygroscopic curves of different electrolytes. One possible way to further examine the role of AWC calculation method is to replace the hygroscopic curves in ISORROPIA by the ones from WRF-Chem.

Figure5&6: it is good to show the different subplots in the same vertical scale or write a note to remind the different scales.

Citation: https://doi.org/10.5194/gmd-2022-62-RC1
RC2:
'Comment on gmd-2022-62', Anonymous Referee #2, 27 Apr 2022
Summary

This paper discusses aerosol pH in China, examining the sensitivity of aerosol pH to non-volatile cations, ammonia (NH3) emissions, chloride (Cl-) emissions, aerosol aqueous phase chemistry, and assumed properties of the aerosol. Quantifying aerosol pH from observations and diagnosing its value in chemical transport models are needed to better understand impacts of aerosol pH on biogeochemical nutrient cycles and generation of reactive oxygen species that can cause aerosol toxicity and adverse health effects. Thus, the paper is an appropriate topic for GMD to publish.

This is a very good paper and should make a substantial impact on understanding processes affecting aerosol pH. The paper is fairly well written, although several spots need to be clarified. My suggestions of needed clarifications are given below.

Major Comments

The authors conclude that non-volatile cations (NVC), ammonia (NH3) emissions, and chloride (Cl-) emissions influence the aerosol pH over different areas of China. However, there is no presentation of the spatial distribution of NVCs, NH3 emissions, or Cl- emissions. It would be useful to see the overlap regions of NVCs and NH3 emissions as well as the main locations for Cl- emissions and SO2 emissions. Please add these maps.

Section 4 tends to repeat what was said in Section 3. Please review Section 4. I suggest writing it in a manner where the reader is reminded of the Section 3 result but spends more text focused on new information and discussion.

Do the authors think the aerosol pH results would differ much with finer horizontal resolution? I could imagine that the grid size of emissions and mesoscale flow patterns influences the range of aerosol pH. One of my comments below brings up this question again in terms of thinking of the average aerosol pH versus the range of aerosol pH during haze events and whether that range would change if the grid spacing were smaller.

Specific Comments

I highly recommend using present tense instead of past tense.

Line 15, in abstract: Why is it critical to estimate aerosol pH accurately in chemical transport models as well as policy development? Please add a phrase.

Section 2.1.1. Does the WRF-Chem simulation include observational nudging so that the meteorology is represented well?

Section 2.1.2. It would be good to add more information on how dust is treated in the model especially in relation to the ions associated with dust? Do iron, magnesium, and/or manganese affect the aerosol pH?

Line 189. I am a little confused between dust and OIN in MOSAIC. I thought in WRF-Chem MOSAIC that dust is part of OIN (other inorganics). Therefore, it is not clear how the speciation profile is implemented when calcium, magnesium, sodium, and potassium are changed. It would also be good to know the speciation profile for the default simulation – please add a sentence in the second paragraph of Section 2.3. Please add a map to show where NVCs are high in concentration over China.

Lines 192-198, please add maps of NH3 and Cl- emissions. It would be informative to explain the emissions sources of ammonia and chloride. How much did Cl- emissions increase?

Section 2.4, It is my understanding to calculate aerosol pH with ISORROPIA constrained by observations that gas-phase NH3 and HNO3 must be known. Could the authors please describe the measurements of NH3 and HNO3, or what values are used for these semi-volatile gases?

Lines 243 and throughout the manuscript. It seems that the number of significant digits is too high for reporting aerosol pH, especially considering the standard deviations that are 20-50% of the actual value. Thus, for Line 243, I suggest reporting 4.2 +/- 2.2 and 5.7 +/- 1.4 for aerosol pH over the Gobi and Taklimakan Deserts, respectively. No need to use “~”.

Lines 265-268, please explain why the doubling of NH3 emissions caused large or small changes in aerosol pH for different regions of China. This explanation would be aided by the maps of NVCs and NH3 emissions.

Line 273, please include a figure showing that Cl- concentration is underestimated compared to observations.

Lines 273-275, deserves more description or explanation since the authors are advocating for future research on chloride emissions.

Line 289. Is the analysis over Beijing averaged over several grid points or is it for a single location?

Section 3.2, For the diurnal variation analysis, how did temperature and relative humidity vary? Is there a correlation between these state variables and aerosol pH? I would imagine that lower T, higher RH, which occur in the early morning, would have more aerosol water, diluting the hydrogen ion concentration.

Line 315, Since there were not NH3 observations made, what is the uncertainty in aerosol pH from ISORROPIA when using the empirical equation? This could be quantified by running ISORROPIA with +/-10% changes in NH3 concentration.

Lines 323-333. This paragraph reports results from Figure 4 but it does not provide any insight. Instead of giving details, I suggest discussing the meaning (or the point) of the results. For example, comparing CTL1 simulation to the default simulation could be written as, “When NVCs are increased, the aerosol pH increases by 0.9 on average with the largest increase occurring during clean periods because …. In contrast, when NH3 emissions are doubled, the aerosol pH increase was smaller (0.4 pH units) compared to CTL1 simulation because ….. With higher NH3 emissions, the pH increased more in the more polluted regions because …..”. Please give explanations for each of the sensitivity cases.

Lines 337-340. Why is aerosol pH sensitive to sulfate production in NCP? Why is the phase state assumption important to TD and GD regions? What are the “influencing factors” to the evolution of aerosol pH in a haze development cycle?

Line 345, please explain more how NVCs (or aerosol composition in general) affect aerosol water amount?

Sections 4.1.1 and 4.1.2 would both benefit from showing maps of the NVCs and NH3 emissions. It would be interesting to see the juxtaposition of these cations.

Line 388, I like Figure 5. Panel a) clearly shows non-linearity within each region. However, is Figure 5b truly non-linear? Eyeballing each region seems to show a linear response but with different slopes between regions. Perhaps a fitting line(s) could be added to the figure.

Lines 405-411, I suggest moving Figures S4 and S5 to the main text since they are discussed at length and provide important support to the conclusion that models need to represent both stable and metastable aerosols.

Sections 4.1.3 and 4.1.4, It seems that producing Figure 5 plots for aerosol phase state and for sulfate production would be interesting. Could panels be added to Figure 5 for these additional simulations?

Line 443, could the authors explain the interfacial chemistry a little bit more? What is it?

Lines 460-465, I think it is important to emphasize that aerosol pH during heavy pollution events remains < 5.0 noting that there is never a time when S(IV) + NO2 formation of S(VI) contributes significantly. Please add! Of course, this is for Beijing region and at dx=36km, which may average out some extreme situations where pH could go higher. When examining individual grid point temporal variations, are there times when pH > 5.0? What might happen if dx=12km (i.e. what might be the impact of the grid spacing on the results here)?

Lines 470-480 or so. The discussion of Figure S8 is very useful. I suggest moving the figure into the main manuscript.

Lines 489-495, it may be beneficial to add a schematic describing what is said in the text here.

Line 499, why does the acid effect prevail over the dilution effect?

Section 4.3 provides useful information, but Pye et al. (2020) already presented these results. This section could be omitted or moved to the supplement.

Section 5. The Conclusions fall short in bringing the results to the greater context (or implications elsewhere). For example, instead of repeating “A priori assumption that aerosols are stable or metastable…”, it could say that across China both stable and metastable state of aerosols exist, thus both states should be represented in regional and global models.

Section 5. The Conclusions did not say anything about the S(IV) + NO2 contribution to S(VI) yet this is part of the Abstract.

Section 5. The Conclusions should also have more text about future needs. Why do we need more high-resolution observations? What is meant by “high resolution”? Do we need higher temporal resolution, or higher spatial resolution, or both? Would it be useful to have measurements throughout the boundary layer? What is the influence of boundary layer mixing on what is being measured? It seems that there is still a lot to be learned about winter-time haze and aerosol pH, and the authors have an opportunity to give their expert opinions on what should be prioritized.

Technical Comments

Line 16, change “reported” to “report”

Line 18, add comma to read as “state assumption, and heterogeneous production”

Line 47, remove “for nowadays”

Lines 99-106, Lines 110-119, use present tense

Line 110, cite Grell et al., 2005 and Fast et al., 2006 when introducing WRF-Chem model

Section 2.3, please use present tense

Line 187, just cite Fig. S3. No need to specify each of the panels.

Line 206, remove “The last” and change “production as which” to “production for which”

Line 268, cite Figure S2.

In the paper, Figure S3 is cited before Figure S2. Please review the text and order of figures.

Line 278, “In particular, PM2.5 pH decreased by 1.9 for TD and 1.1 for GD, reducing aerosol pH values to 4.8 and 4.0, respectively, whereas the metastable state assumption had little impact ….”

Line 293, use present tense (“were” to “are”)

Line 303, remove “On the other hand”. I’m not sure what is being contrasted.

Line 315, remove “so”

Line 315, This sentence about NH3 observations should be in section 2.4. Please add the empirical equation to the supplemental text (or in the main text).

Section 3.2, Figure 4b is not discussed.

Lines 337, reword the sentence removing “In addition”. Maybe “For NCP where severe and frequent haze events occur, PM2.5 pH is very sensitive to the magnitude ….”

Lines 347-354 are stated in the Results section and do not need to be repeated here.

Line 362, change “minimum” to “minimal”

Line 364, change “may” to “can”. We know aerosol thermodynamics makes this sentence true. Same thing with Line 366 (“may” to “can”). You could also support these statements with supplement information using ISORROPIA calculations.

Line 376, no need for both “In addition” and “also” in the same sentence. This occurs many times in the paper. Please proofread and try avoid using “in addition” and “also”.

Line 378, change “But” to “However,”

Line 416, change to “all size bins”

Line 421, change “might be” to “is” as you just spent two paragraphs showing the consequence of this assumption.

Line 431, could you be more explicit about “differ by regions”? Where would the sulfate production be buffered?

Line 438, the authors may want to note that H2O2 oxidation may be small because of low OH during the time of year investigated for the study.

Line 450, please state why CTL3meta was chosen for further analysis instead of any of the other simulations.

Line 460, change to “under heavy pollution events …. was 3.6 +/- 0.5”.

Line 485, delete “also”

Line 486, change to “high on more polluted”

Line 512, change to “both models using the same”

Table 1, add vertical domain and grid (i.e., number of vertical levels, how many levels in the boundary layer). Also add information about observational nudging of the meteorology.

Figure 1a. The six sub-regions are hard to read. Try using gray for the geography and a darker color for the acronyms.

Figure 2, change CLT to CTL (check all figures for this misspelling)

Figure 3 panels are in a different order than Figure 2

Figure 4b, could WRF RH line be black (to be consistent with observations) and maybe dashed.

Figure 7, are these results for all the simulated days or for just the polluted days?

Table S2, change to “ranges and mean” because of the order of the columns.

Table S3, where are the “decreasing regions”? Be more descriptive. Also add the time period averaged over for these results.

Figure S3a. Should the observations be plotted as markers for when the measurements are taken (like in S3b)? A line for the observations could be misleading.

Figure S3. The last two sentences are not clear. Why are they included in the figure caption?

Figure S4. Is it surface relative humidity or RH at 2m?
Citation: https://doi.org/10.5194/gmd-2022-62-RC2

Xueyin Ruan et al.

Supplement

https://doi.org/10.5194/gmd-2022-62-supplement

Model code and software

xyruan/wrfchem_ustc4.0: wrfchem_ustc 4.0 including heterogeneous sulfate production in aerosol water and modified cation speciation profiles Xueyin Ruan, Chun Zhao, Rahul A. Zaveri, Pengzhen He, Xinming Wang, Jingyuan Shao, and Lei Geng https://doi.org/10.5281/zenodo.6359417

WRF Source Codes and Graphics Software Downloads WRF Users Page http://www2.mmm.ucar.edu/wrf/users/download/get_source.html

Xueyin Ruan et al.

Viewed

Total article views: 367 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
301	55	11	367	16	2	4

HTML: 301
PDF: 55
XML: 11
Total: 367
Supplement: 16
BibTeX: 2
EndNote: 4

Views and downloads (calculated since 21 Mar 2022)

Month	HTML	PDF	XML	Total
Mar 2022	149	24	4	177
Apr 2022	110	29	6	145
May 2022	42	2	1	45

Cumulative views and downloads (calculated since 21 Mar 2022)

Month	HTML	PDF	XML	Total
Mar 2022	149	24	4	177
Apr 2022	110	29	6	145
May 2022	42	2	1	45

Viewed (geographical distribution)

Total article views: 311 (including HTML, PDF, and XML) Thereof 311 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 20 May 2022

Short summary

Accurate prediction of aerosol pH in chemical transport models is essential to aerosol modeling. This study examines the performance of Weather Research and Forecasting model coupled with Chemistry (WRF-Chem) on aerosol pH predictions and the sensitivities to emissions of nonvolatile cations and NH₃, aerosol phase state assumption and heterogeneous sulfate production. Temporal evolution of aerosol pH during haze cycles in Beijing and the driving factors are also presented and discussed.


Total:	0
HTML:	0
PDF:	0
XML:	0