Preprints
https://doi.org/10.5194/egusphere-2026-1730
© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/egusphere-2026-1730
© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
04 May 2026
| 04 May 2026
Status: this preprint is open for discussion and under review for Atmospheric Chemistry and Physics (ACP).
Impacts of aerosols on the tornado potential: A case study in Yangtze River Delta, China
Abstract. Extensive observational and modeling studies have demonstrated that aerosols can modify extreme rainfall and convective storms. Their impacts on the genesis and development of tornadoes remain largely unexplored. By incorporating data assimilation into WRF-Chem simulations, we successfully simulate the whole life cycle of a supercell tornado and demonstrate the roles of anthropogenic aerosols. It is found that aerosols can enhance tornado potential, quantified here by the Significant Tornado Parameter, and affect storm motion, precipitation evolution, and cold-pool structure chiefly through two mechanisms. First, aerosols enhance condensational heating within the 0.3–1 km layer. Second, aerosols reduce near-surface evaporative cooling within the low-level updraft core by shifting it away from regions of strong rain evaporation. Together, these thermodynamic effects increase heating and thermal buoyancy, accelerating the low-level updraft. The aerosol-caused strengthening of the updraft enhances low-level convergence and deepens storm-relative inflow, leading to increased ingestion of streamwise vorticity in the 0–1 km layer, which dominates the enhancement of the tornado potential. This study gains new insights into the thermodynamic and dynamical pathways through which aerosols can influence extreme weather.
How to cite. Wang, R., Fan, T., Li, Z., and Guo, J.: Impacts of aerosols on the tornado potential: A case study in Yangtze River Delta, China, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2026-1730, 2026.
Competing interests: One of the authors, Zhanqing Li, is a member of the editorial board of Atmospheric Chemistry and Physics. The authors declare that they have no other competing interests.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.Download & links
- Preprint
(17002 KB) - Metadata XML
-
Supplement
(6726 KB) - BibTeX
- EndNote
Status: open (until 09 Jul 2026)
Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor
| : Report abuse
- CC1: 'Comment on egusphere-2026-1730', John M Peters, 13 May 2026 reply
-
CC2: 'Comment on egusphere-2026-1730', Paul Markowski, 13 May 2026
reply
I would like to see a skew T diagram of the environment. The reader also isn’t told explicitly what the environmental LCL, SRH, or CAPE are (among other things about the environment that are calculated from a sounding or hodograph). From Fig. 5 (which shows the various contributions to STP) and eqn (1) (the eqn for STP), it looks like the MLLCL contributes approximately 0.6–0.7 to STP, which would imply an MLLCL of 1 km. Given this, how can there be so much condensational heating below 1 km? Is it because the storm has a wall cloud that extends much closer to the ground than the predicted cloud base height? I am skeptical that a supercell in a simulation with dx = 1 km would have such a wall cloud though (a substantial wall cloud is unlikely to be resolved in my experience). All of this condensation below the cloud base (at least the cloud base one can infer from Fig. 5) needs to be explained. I admit to being unfamiliar with WRF, but in a few places it’s mentioned that the thermodynamic budget isn’t reconciled. Would that not suggest a bug, and if the budget has known problems, can we trust the budgets? The claim of condensational heating in the 300-1000 m layer should raise eyebrows if the environmental LCL is indeed at roughly 1 km.It is also argued that aerosols enhance tornado likelihood by increasing STP. Why not also look at the simulated supercells themselves in order to judge tornado potential (e.g., max near-surface zeta as a proxy for tornado threat)? It would seem better to look at the rotational properties of the supercells themselves given that there’s more to tornado formation within a supercell than simply the environmental STP. Regarding the calculations of STP, I have trouble following the methodology described in section 2.4, but it seems to be saying that STP values were averaged within an area that is within a grid cell of 30 dBZ reflectivity. That’s awfully close to the storm, and not really in the region previous studies typically have evaluated storm environments. It’s unclear to me if this averaging region lies in the storm inflow.Citation: https://doi.org/
10.5194/egusphere-2026-1730-CC2 -
RC1: 'Comment on egusphere-2026-1730', Anonymous Referee #1, 02 Jun 2026
reply
The authors attempted to assess the anthropogenic aerosol effects on tornado potential by conducting WRF-Chem sensitivity modeling work. They conducted a relatively comprehensive analysis to identify and quantify the aerosol-induced thermodynamic and dynamic effects to interpret tornado potential response to aerosols. However, I have some concerns regarding correctly isolating the real aerosol effect and accuracy of reporting and interpreting some results.
- One of my concerns is about the aerosol effect isolation using the sensitivity modeling approach. When the frequent assimilation of AWS and radar observations improve storm representation, it may distort or even mask the aerosol effects on thermodynamic and dynamics. Also, because the spin-up of chemistry was prior to the assimilation, the aerosol effect evaluated during the analysis period seems also from the continued influence during data assimilation; the interpretation of (ALL–NoAero) differences as aerosol-driven becomes less clear during the free forecast period after assimilation. What are the initial conditions right before the analysis period for both meteorology and aerosol fields? Are they same between ALL and NoAero simulations? If not, the authors should be careful when interpreting the ALL–NoAero differences as aerosol effects during the analysis period.
- Line 125: the statement of “differences between experiments arise solely from aerosol perturbations” is sort of too strong. The authors need to tune down the tone according to the results of the further analysis mentioned above. Uncertainty remains due to nonlinear model response and data assimilation constraints, even with identical assimilation settings.
- Lines 229: the authors reported that the aerosol impacts identified in the current experiment settings might be much less significant than the real situation due to the largely underestimated aerosol concentration in ALL (Fig. 3d). It would be helpful also to report the aerosol field evolutions in both ALL and NoAero, or the differences between them, to show how considerable the aerosol perturbations of NoAero from ALL are before and during the tornadic event.
- Were the same 70 members selected for both ALL and NoAero experiments for analysis? It’s also not clear how the selection criteria (e.g., the selection metric, variables used, evaluation period, etc.) was defined. Was the aerosol field also included in the selection criteria?
- Lines 291-292: The net heating by aerosols appears minor in Fig. 9a because of the almost right offsets b/w microphysical heating and advective cooling. It does not make sense linking such little change in net heating to the enhanced buoyancy caused by aerosol effects. In addition, what are the mechanisms leading to significant alterations in advective cooling by aerosols?
- Lines 298-299: Hard to tell the aerosol signature in the displacement of updraft core from rain evaporation region from Fig. 10. To conduct a comparison of the raindrop mass vertical profiles (averaged over updraft cores) of the two simulation sets (or their difference) can provide more direct support for this argument. In addition, statement in Line 299 is redundant relative to Line 298, which can be removed.
- Please detail the aerosol-related schemes (both physics and chemistry) adopted in the model in this study for a clear context if comparing with other similar studies. Also provide cloud microphysics and radiation scheme info.
- Section 3.4: I feel that the proposed “new mechanism” for aerosol condensational invigoration is sort of overstated, given that bunch of previous studies have already elaborate similar aerosol condensational invigoration at cloud microphysical process. There are not too many new insights based on the current results in this study.
Citation: https://doi.org/10.5194/egusphere-2026-1730-RC1
Viewed
Total article views: 386 (including HTML, PDF, and XML)
Cumulative views and downloads
(calculated since 04 May 2026)
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 325 | 34 | 27 | 386 | 22 | 17 | 16 |
- HTML: 325
- PDF: 34
- XML: 27
- Total: 386
- Supplement: 22
- BibTeX: 17
- EndNote: 16
Viewed (geographical distribution)
Total article views: 379 (including HTML, PDF, and XML)
Thereof 379 with geography defined
and 0 with unknown origin.
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
1
Latest update: 13 Jun 2026
Rumo Wang
Faculty of Geographical Science, Beijing Normal University, Beijing, 100875, China
Faculty of Geographical Science, Beijing Normal University, Beijing, 100875, China
Zhanqing Li
Department of Atmospheric and Oceanic Science, Earth System Science Interdisciplinary Center, University of Maryland, USA
Jianping Guo
State Key Laboratory of Severe Weather Meteorological Science and Technology & Specialized Meteorological Support Technology Research Center, Chinese Academy of Meteorological Sciences, Beijing, 100081, China
Short summary
This study investigated whether aerosols from anthropogenic emissions can make storm environment more favorable for producing tornados. We numerically modeled a real storm in the populated eastern China and found that aerosols can warm the low-level air and strengthen the updraft, allowing the storm to more effectively ingest environmental vorticity, and thus favoring the production of tornados. This suggests that air pollution may increase severe weather risk in densely populated regions.
This study investigated whether aerosols from anthropogenic emissions can make storm environment...
This article attempts to associate aerosols with tornado potential in supercells. I have found numerous errors, inconsistencies, incorrectly cited articles, and misrepresentations of past research. My comments are listed below:
1) P2, L30: The authors state that research indicates that aerosols can invigorate microphysical heating. However, there are numerous articles that argue that aerosol invigoration of microphysical heating is unfeasible. For instance:
Igel and Van den Heever (2021): https://doi.org/10.1029/2021GL093804
Grabowski and Morrison (2020): https://doi.org/10.1175/JAS-D-20-0012.1
Romps et al. (2023): https://doi.org/10.1029/2022GL100409
Peters et al. (2023): https://doi.org/10.1029/2023GL103314
2) P2, L30: The authors state that condensational heating is necessary for tornadogenesis. However, there is no literary basis this connection that I am aware of. The authors need to provide a citation.
3) P3, L85: A radar-observed velocity couplet does not necessarily indicate that there is near-ground rotation. This only shows rotation at the elevation of the radar beam.
4) P8, L 145: The method here for estimating storm motion (C) is outdated. The standard among the forecasting and research community is the method introduced by Bunkers (2000; https://doi.org/10.1175/1520-0434(2000)015<0061:PSMUAN>2.0.CO;2), to be used when the storm motion is not known. However, the authors have a simulated storm, they should use the simulated storm motion.
5) Equations 6a-6e: This decomposition is not useful because it is not Galilean invariant (see Davies Jones 2002; https://doi.org/10.1175/1520-0469(2003)059<3178:LANPOS>2.0.CO;2).
6) Eq. 6b is written incorrectly: This equation should contain squared first derivatives (du/dx)^2 + (dv/dy)^2 + (dw/dz)^2, not second derivatives. The way the equation is currently written, the units on the left-hand-side do not equal those on the right-hand-side.
7) P10, L215: These are the incorrect boundary conditions for the pressure decomposition. The authors use Dirichlet conditions (p’=0) on all boundaries. However, the lower boundary for dynamic pressure needs Neumann conditions (dp’/dz = 0), and the lower boundary for buoyancy pressure is (dp_b/dz = -rho B). This error invalidates all analysis using the pressure decomposition, especially near the ground. See (see Davies Jones 2002; https://doi.org/10.1175/1520-0469(2003)059<3178:LANPOS>2.0.CO;2) and numerous other articles.
8) P11, L230: This statement about the model underestimating aerosol impacts is purely speculative. There is no citation or evidence to provide a basis for this statement.
9) Section 3.2: The authors should focus on the near-ground characteristics of the simulated storm (such as near-ground rotation), rather than STP, to assess aerosol impacts on tornadoes. Otherwise the conclusions are tenuous at best.
10) P14, L260: The relationship between low-level inflow, low-level mesocyclones, and tornadoes is poorly understood (see Peters et al. (2023); https://doi.org/10.1175/JAS-D-22-0114.1). Hence, increases in inflow do not necessarily influence tornado potential.
11) P17, L295: The authors attribute differences in condensational heating to aerosol effects. However, there are a million different reasons why there could be differences in condensational heating between the simulations, ranging from stochastic differences in storm behavior to differences in meteorological characteristics within the storm environments. Hence, there is no causal relationship between aerosols and condensational heating demonstrated here.
12) P18, L300: Aerosol-driven enhancement in condensational heating is not “well-known.” In fact, it is a hotly debated topic with numerous papers suggesting that it is implausible. The authors have conveniently neglected to cite or discuss these papers, as I mentioned earlier.
13) P20, L310: Why are we even talking about condensational heating below 0.3 km? At these levels, parcels are likely to be subsaturated so condensational heating is a moot point.
General conclusions: The conclude that aerosols enhance low-level updraft buoyancy, which drives more inflow, increases the STP, and therefore enhances tornado potential. However, they have not at all demonstrated a causal relationship between aerosols and microphysical heating. The differences in storm behavior between the two simulations could be attributable to a multitude of other factors. This lack of causality is especially concerning given the large body of research (which is ignored in this paper) that has shown that substantial enhancement of condensational heating by aerosols is implausible. Hence, this conclusion is unlikely to be true. It is almost certain that meteorological factors other than those emphasized in this article are responsible for the differences seen between the two simulations.