Idealized Simulations of a Supercell Interacting with an Urban Area

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

I could not find a reference in the text describing Table 1 and its interpretation.  The data therein suggest very small, possibly insignificant differences in the values upwind and downwind of the city, and for that reason this table could probably be removed from the manuscript without loss of important information.

 A relatively minor annoyance is that some of the figures are displaced rather distant from where they are discussed in the text (e.g., Figs. 4-6), which causes paging back and forth, something that can easily be avoided by inserting the figure immediately after the place it is first mentioned in the text.

Please specifically identify in the figure caption what is displayed in each panel in Fig. 6, even though the units suggest which panels are SRH and which are mixing ratio.  Also, a scientific explanation is needed in the text for why the SRH (which uses the storm-relative winds) is lessened over the city while the UH (the product of the vertical motion and vertical vorticity) is enhanced by the simplified city. 

Likewise, Fig. 6 does not display enough information to the reader to confirm the assertion that "the ribbon of enhanced 0-2 km SRH on the north side of the simplified city is further amplified as it interacts with the forward-flank gust front."  Adding a figure that depicts the cold pool and associated gust front would help.  While relevant kinds of analyses are depicted in Fig. 7 with regards to other features, they do not directly address this aspect (quoted text) here.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript under review presents a series of idealized simulations that vary the perturbation skin surface temperature and roughness length of a 10-km radius city to assess the influence of the city on properties of simulated supercells. The subject is novel, worthy of exploration, and CM1 is well suited to this type of assessment. My primary concerns are rooted in several of the choices in methodology, which I detail below. Based on these, I recommend acceptance pending major revision, so that the authors have the opportunity to either justify these choices or run different simulations that account for them.

1. Environmental wind profile: The authors justify use of no method to maintain a steady wind profile through analysis of SRH over the first 45 minutes of the simulation. Any decrease in shear through vertical momentum flux will depend on the depth and nature of thermals, which is not discussed in the paper (see comment 6). It is entirely feasible that these values did not decrease substantially by t = 45 min, but did between then and t ~ 90 min, when the cold pool began to interact with the city. At a minimum, this later time needs to be included in Table 2.
2. Laminar inflow: Simulations that explicitly resolve boundary layer convection are known to have laminar inflow at upstream lateral boundaries, with eddy deficient flow extending well into the model domain (see Boyer and Keeler, 2022; https://doi.org/10.1175/JAMC-D-22-0017.1). Given that this is a boundary layer study, it is critical that the authors demonstrate that this issue has been mitigated somehow. I am concerned that this is not the case, since no method to address this has been discussed (e.g., eddy seeding, as available in recent versions of CM1).
3. Radiation: A separate influence on the storm evolution would be anvil shading, which is not accounted for in these simulations. Discussion of the relative magnitude of cooling from anvil shading would add helpful context, relative to the magnitude of UHI thermal perturbation used here. For example, the magnitude of observed anvil shading discussed in Frame and Markowski (2013) is of sufficient magnitude to substantially offset the UHI 5 K warm perturbation used herein.
4. Magnitude of UHI thermal perturbation: What is the justification for the use of a 5K skin surface thermal perturbation in the urban area? Is this a typical value? Can the authors include references demonstrating this based on prior observational work?
5. Horizontal grid spacing: Are the authors’ result sensitive to the horizontal grid spacing chosen? Prior work has shown that use of horizontal grid spacing greater than 200 m can result in substantial error in representation of surface fluxes (Grand and van den Heever, 2016).
6. Spin up of relevant BL circulations: In its current form, the manuscript lacks discussion of spin-up of relevant boundary layer structures, such as cellular convection, horizontal convective rolls, or (surprisingly) the UHI circulation. Were these structures discernable and in a steady-state form by the time the supercell interacted with the urban area? Evidence of this is needed to convince readers that the timing of this interaction is appropriate.

References

Frame, J. and P. Markowski, 2013: Dynamical Influences of Anvil Shading on Simulated Supercell Thunderstorms. Mon. Wea. Rev., 141, 2802-2820, https://doi.org/10.1175/MWR-D-12-00146.1.

Frame, J., and P. Markowski, 2013: Dynamical Influences of Anvil Shading on Simulated Supercell Thunderstorms. Mon. Wea. Rev., 141, 2802–2820, https://doi.org/10.1175/MWR-D-12-00146.1.

Grant, L. D., and S. C. van den Heever, 2016: Cold pool dissipation. J. Geophys. Res. Atmos., 121, 762 1138–1155, doi:10.1038/175238c0.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Idealized Simulations of a Supercell Interacting with an Urban Area 

1. The authors choose a circular city methodology. Any thoughts concerning change in the size and shape of a city in ideal scenarios? Studies have been conducted in real case simulations worldwide considering the size and shape factors.
2. The city simulations used a bulk surface roughness and thus the parcel movement can be like an orographic lifting scenario. Any thoughts on that? Can a variable surface roughness be performed to represent ideal city roughness at such a larger scale?
3. Is temporal resolution playing any role?? Why there is similar behavior till t=90 for all simulations?
4. Discussions concerning boundary layer variations and boundary layer parameterizations are lacking in the second paragraph of the results. The planetary boundary layer plays an important role in urban interaction studies. Authors are suggested to add exclusive discussion on this.
5. Figure 4 caption: The 0–1 km UH values are filtered based on a 2–5 km UH threshold of 150 m2 s􀀀2. How the filtration is explained. The color bar shows values from 20 to 200 for the same creating confusion.
6. Line 148-150. What could be the abrupt peak in the case of control simulation? Also noticed in time series plots in Figure 5a.
7. Mention what represents each subplot in Figure 6 captions.
8. The effect of the city is also noticed as it bifurcates the storm with enhanced SRH values seen around the circular city in Figure 6. This can also be discussed as a part of the conclusion.
9. What is the role of the city in storm movement? How much time does the storm take to move from the start point to the end point of the city? Is there any time difference noticed between the control and city simulation regarding the time taken by the storm to pass?
10. Any specific reason to use the CM1 version 18, i.e. released almost 8 years ago, while much newer versions released after that are already available (Version 21, Released in 2022, Version 20. Released in 2021.)?
11. What are the major differences in CM1 w.r.t the WRF model? Most of the modern urban impact studies adopt WRF to be a suitable model for their investigation. Do you think WRF ideal simulations can be better supporting evidence in this regard?
12. A schematic diagram or representation of the simulation scenario including a circular city, storm starting location, storm movement direction, etc. would be interesting to imagine the results while reading it.
13. The current conclusion section fails to produce the innovativeness of the study. The authors suggested re-write the conclusion keeping in mind the review comments and objectives defined in the introduction section.

Overall, the manuscript is very well written, and the analysis provides evidence for the urban build-up of storm structure. However, most of the conclusions are well observed using real data simulations using WRF and observational evidence. Authors are suggested to highlight What additional or new results or new understanding is obtained from this ideal storm-resolving simulation.

As mentioned above, research groups worldwide have already investigated the city's impact on thunderstorms in detail using real-world simulations. The meta-analysis by Niyogi et.al group and Panda et.al group has conducted very high-resolution city-scale simulations that suggested the impact of the urban area on various aspects of thunderstorms.

Can idealized simulations put some light on the below scenarios:

1. Impacts of the storm that is already buildup before passing through a city than the storm gradually building up while passing over a city.
2. In the two scenarios mentioned above the rainfall over the city and downward of the city probably vary significantly. The current result shows almost the same rainfall over the city though it is enhanced downward in the city.

Please discuss these aspects if possible.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

I thank the authors for their efforts to address my earlier concerns. My primary remaining concern is related to the realism of their simulated boundary layer structure and the implications that has on their results. While the authors include perturbations at model initialization, Boyer and Keeler (2022) showed that these initial perturbations alone are not sufficient to prevent unrealistic, laminar flow in the boundary layer in idealized simulations. A simple fix that the authors could implement is use of eddy recycling, available starting in version 21.0 of CM1. I request that the authors either run new simulations using eddy recycling (preferred), or demonstrate that the boundary layer is fully "spun up" by the time inflow from the upstream lateral boundary reaches the city. This could be demonstrated through an analysis similar to that shown in Figure 17 of Boyer and Keeler.

Boyer, C. H., and J. M. Keeler, 2022: Evaluation and Improvement of an Inflow-Nudging Technique for Idealized Simulations of Convective Boundary Layers. J. Appl. Meteor. Climatol., 61, 1843–1860, https://doi.org/10.1175/JAMC-D-22-0017.1.

Author Response

Please see attachment

Author Response File: Author Response.pdf

Round 3

Reviewer 2 Report

Comments and Suggestions for Authors

I thank the authors for their efforts to address my concerns.