the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 26 Aug 2025
The Entrainment of Air from Rainy Surface Regions and its Implications for Bioaerosol Transport in Three Deep Convective Storm Morphologies
Abstract. The rain produced by thunderstorms has been observed to coincide spatially and temporally with enhanced near-surface concentrations of warm-temperature ice nucleating particles (INPs) of biological origin. However, the air in rainy regions is evaporatively cooled and negatively buoyant, and so it is unclear if it is entrained into its parent storms. Despite bioaerosols being highly ice-nucleation active, the microphysical influence that rain-aerosolized bioaerosols exert on storm processes is therefore not well-understood. We use the RAMS cloud-resolving model to simulate high-resolution archetypal representations of three deep convective storm morphologies: isolated deep convection, a squall line, and a supercell. We measure the degree of entrainment of rainy and non-rainy surface air into its parent storm using passive tracers, as well as calculating measures of each storm’s characteristics that influence the timing and degree of this entrainment. We find different degrees of entrainment between storm morphologies and between rainy and non-rainy surface air, with the squall line and supercell entraining significantly more rainy air than the isolated convective storm for all but the lightest rain. These differences owe to variation between the storms in their degrees of entrainment of surface air, their proportions of entrained surface air that originate in rainy regions, and their amount of rain produced per updraft mass. This study finds a specific and previously unrecognized source of air potentially containing highly ice-active aerosols which is entrained to varying degrees in different convective storm morphologies, and which is likely to exert different microphysical impacts on each type of storm.
- Preprint
(5694 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
EC1: 'Editor Comment on egusphere-2025-2968', Johannes Quaas, 15 Sep 2025
-
AC1: 'Reply on EC1', Charles Davis, 16 Sep 2025
Dear Johannes, thank you for passing along this comment. We are looking into it.
Kind regards,
Charles
Citation: https://doi.org/10.5194/egusphere-2025-2968-AC1 -
AC2: 'Reply on AC1', Charles Davis, 20 Nov 2025
Dear Johannes and reviewer, we have compiled a response to the comment in the attached document. We look forward to your response.
Kind regards,
Charles Davis
-
RC2: 'Comment on egusphere-2025-2968', Anonymous Referee #2, 03 Dec 2025
The manuscript concerns pathways of tracers in atmospheric deep convection. The topic is scientifically significant. However, the presentation quality is low. For example, some of the terms are incorrect or misleading, definitions are not rigorous, use of different words or phrases for the same quantity makes reading challenging. Figures are not properly discussed. Due to to the low presentation quality, scientific quality is difficult to assess.
Due to critical problems in the quality of presentation, a thorough review is not possible. I recommend that this manuscript is rejected.
In the following, there is a more detailed description of some of the problems. Note that the order does not indicate the seriousness of the problem.
- Line 177-180 Do you use rainfall thresholds or rainfall categories: 1 mm/hr – 5 mm/hr, 5 mm/hr – 10 mm/h and so on. Otherwise, tracer species for the lightest rain is released for all rainfall rates above the value of 1 mm/hr.
- New Table 3 (old table 2). In the following, the four terms in the table are called A, B, C, and D. The “physical question answered” for A is misleading. Based on your new table 2, A is simply the mean FS tracer mixing ratio in the updraft. It is not how much air the updraft entrains from the surface. For the same amount of air entrained from the surface layer, A would be larger for a later time than for an earlier time since the total emitted FS tracer increases linearly with time (Fig. 2a). Also, lateral entrainment affects the value of A. Therefore, there are many problems in the “physical question answered”. Overall, using A in comparisons is problematic since time from the beginning of the simulation directly affects the amount of FS tracers. Only on lines 335-339 do you mention that A is also affected by how long the simulations has been going on. Also, B is affected by the same problem. If you compare two different storm types, differences in their A and B can be due to how long emission has been going on. I recommend reconsidering whether to use A and B in your study at all. At least you should have discussed this problem when you first introduce A and B.
- What is the meaning of “total”? In some places it seems to relate to summing from time 0 to present. In other places, its meaning is not clear. Instead of using a word which can be understood in many ways, please be precise and explicit.
- Term D and “Rain production efficiency”. The physical question answered is wrong and the Rain production efficiency is a misleading term. Based on the new table, D is the total number of RS tracers emitted until “present” divided by “current” updraft mass. Why do you divide by the “current” updraft mass, or by any mass whatsoever? Note that if the updraft mass drops to one tenth of what it was before, term D would increase tenfold. Would this mean that rain production suddenly increases and that the storm would produce much more rain (per updraft)? No. This same comment applies to the whole section 3.2.4.
- In new Table 2, the total RS tracer emitted is defined as a summation from time 0 to present. Therefore, term D, as mentioned above, is a summation from time zero to present divided by an instantaneous value. However, on line 406 you write that the rain production efficiency (term D) is the “instantaneous rainfall rate to the instantaneous updraft mass”. Moreover, the numerator in D is not “rainfall rate” but “# of RS tracers emitted”. Those two things are not the same.
- Lines 314- 315. “The squall line and supercell achieve similar maximum and end-of-simulation mixing ratios to each other for the 1, 5, and 10 mm hr-1 tracers (Fig 7 (k-l)). “ These figures are missing from the manuscript.
- On line 407 you mention that the large values of D are due to the storm continuing to rain after the updraft begins to decay “rather than it producing an exceptionally large amount of rain per updraft mass.” Here you admit that the “physical question answered” in (new) Table 3 is wrong. Why do you include D in your study at all?
- Section 3.2.3 RS tracer fraction. One storm can have larger value for the RS tracer fraction than another just because it entrains less air from the surface with no precipitation and more from above the surface. What is the reason for not simply discussing the mean in-cloud RS tracer mixing ratios instead? Why do you only show them for the lowest rain threshold emission in Fig. 7?
- Apparently, in term C the total RS tracer emitted is the sum of that emitted up to present time. However, in term D the same name is used for instantaneous value which became clear on lines 406 (see point 5).
- Klemp-Wilhelmson sounding is very moist. Discuss its effect on the results.
- Figures 6-9 are not discussed properly where they are introduced. For example, only figures 6 (j-l), 7(a and c), 8 (a-c), and 9 c are referred to between lines 296- 350. Besides, are all these figures necessary in the manuscript? See comments above.
- When you describe a figure, make sure to also write the figure number and letter.
- Be precise: If you discuss RS tracers in cloud, do not write “RS tracers”.
- Do not write per-updraft if you mean per-updraft mass.
- Use always the same name for the same quantity, i.e., do not use synonyms.
- In Figure 3, colors for rainfall rates are missing.
- Figure 3: what are the low-level storm features that are shown?
- New table 2: The unit of the rain sourced tracer mixing ratio should be #/kg and not #/m3.
- New table 2: Total fixed source tracer entrained and total rain sourced tracer entrained should be without “total” since “total” is used for the following two quantities for which a summing is made from time 0 to present. See also major point 3.
- New table 2. Instead of “Total fixed source tracer entrained and total rain sourced tracer entrained write “(number of) fixed source tracers in updrafts” and “(number of) rain sourced tracers in updrafts” since this is what they are.
- According to the new table 2 it seems that tracers for weakest rain threshold is emitted for all rain categories above 1 mm/hr. Is it so?
- Times in Figure 4 and corresponding times in the text do not match.
- Table 1 “westerly” should be added to the squall line shear description.
- You use the term “Total rain-sourced tracer entrained” and “Total fixed sourced tracer entrained”. It is more precise to call them “rain-sourced and fixed-sourced tracers in updrafts” as this is what they are according to the new table 2.
- Also lines 274-279 should be changed according to comments above.
- line 395: proportion of what?
- line 406: it is not instantaneous rainfall rate. It is instantaneous number of RS tracer emitted per rainfall category.
- lines 489 “size of the updraft” should be changed to “mass of the updraft”.
Citation: https://doi.org/10.5194/egusphere-2025-2968-RC2
-
RC2: 'Comment on egusphere-2025-2968', Anonymous Referee #2, 03 Dec 2025
-
AC2: 'Reply on AC1', Charles Davis, 20 Nov 2025
-
AC1: 'Reply on EC1', Charles Davis, 16 Sep 2025
-
RC1: 'Comment on egusphere-2025-2968', Anonymous Referee #1, 17 Sep 2025
This manuscript uses the high-resolution RAMS cloud-resolving model to compare three types of deep convection (isolated deep convection, a squall line, and a supercell) in terms of how they entrain and transport “rain-sourced near-surface air.” Overall, the topic is novel, the model/diagnostic design is sound, and the results are interesting with potential implications for cloud–aerosol interactions. However, several important issues need to be addressed before the paper can be considered for publication.
Major Issues
- The “rain-sourced” tracer is purely passive and only subject to advection and diffusion. It does not include wet deposition or particle-specific processes. Although the authors briefly acknowledge this limitation, there is no quantitative estimate or sensitivity test to assess how wet scavenging might reduce the actual amount of rain-sourced aerosol that could be lofted.
- Because the tracers ignore wet scavenging, the only real difference between “rain-sourced” and “fixed-source” tracers is whether they are released during rainfall or not. In essence, the study shows the effect of storm winds on air masses, rather than the specific role of rainfall. As such, the link to rain-induced aerosolization (and especially to bioaerosols/INPs) feels somewhat tenuous, and the extrapolation to microphysical or health impacts is weak in its current form.
- The main conclusion is that, if rain-induced bioaerosols exist, storms entrain them to different degrees. However, the paper does not show how much of the rain-sourced tracer actually reaches the upper troposphere (e.g., 8–12 km). This is critical for evaluating possible impacts on ice nucleation near cloud tops or for long-range transport. Prior studies have specifically used “lofting to 10 km” as a benchmark, but the manuscript provides no such metric.
Detail Issues
- The captions of Figures 7–9 should clearly state variable units, normalization methods, and any temporal smoothing or averaging. e.g., Fig. 9: maximum updraft at 5 km AGL smoothed over 15 min. Currently, some details appear only in the text, which makes the figures harder to interpret.
- Model data are saved every 5 minutes (Table 1). Since some entrainment/updraft processes may evolve on shorter timescales, please explain whether this temporal resolution is sufficient for your analysis.
- The y boundary of Squall line is cyclic, while the other cases are radical. Please discuss whether this introduces differences and why different boundaries are adopted.
- The assumption that aerosol release occurs once rainfall intensity passes certain thresholds should be supported by references.
- Can you provide at least a rough estimate of how the tracer concentrations could map onto INP enhancements, e.g., within the DeMott (2010) framework, and what impact this might have on ice nucleation rates?
- Table 1 units: The initial aerosol profile is listed with units of “mg⁻¹,” which seems inconsistent. Please correct the unit and confirm the magnitude.
Citation: https://doi.org/10.5194/egusphere-2025-2968-RC1 - AC3: 'Reply on RC1', Charles Davis, 20 Nov 2025
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 2,013 | 50 | 28 | 2,091 | 27 | 37 |
- HTML: 2,013
- PDF: 50
- XML: 28
- Total: 2,091
- BibTeX: 27
- EndNote: 37
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Dear authors,
I was contacted by one reviewer who has an important comment that hampers review. Please address the comment in the Discussion as soon as possible.
Best regards
This manuscript presents interesting research on pathways of tracers ("biological aerosol particles") in atmospheric deep convection. However, section 3.2.1 has several obscure definitions starting from equation 1. Also, some terms, physical question answered and names in table 2 contradict each other. Could you please revise this whole section. I suggest that you first explain rigorously step by step how each physical quantity discussed in the paper was calculated in the model. Second, please make sure that the names and explanations are not misleading. The rest of the paper needs some editing accordingly.