the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Adverse impact of terrain steepness on thermally-driven initiation of orographic convection
Matthias Göbel
Stefano Serafin
Mathias Walter Rotach
Abstract. Diurnal mountain winds precondition the environment for deep moist convection through horizontal and vertical transport of heat and moisture. They also play a key role in convection initiation, especially in strongly inhibited environments, by lifting air parcels above the level of free convection. Despite its relevance, the impact of these thermally-driven circulations on convection initiation has yet to be examined systematically. Using idealized large-eddy simulations with the WRF model, we study the effect of cross-valley circulations on convection initiation under synoptically undisturbed and convectively inhibited conditions, considering quasi-2D mountain ranges of different heights and widths. In particular, we contrast convection initiation over relatively steep mountains (20 % average slope) and moderately steep ones (10 %). One distinctive finding is that, under identical environmental conditions, relatively steep mountain ranges lead to a delayed onset and lower intensity of deep moist convection, although they cause stronger thermal updrafts at ridge tops. This finding cannot be explained considering the temporal evolution of convective indices, such as convective inhibition and convective available potential energy. Analysis of the ridgetop moisture budget reveals the competing effects of moisture advection by the mean thermally-driven circulation and turbulent moisture transport. In general, at mountaintops, the divergence of the turbulent moisture flux offsets the convergence of the advective moisture flux almost entirely. The weaker total moistening found over steep mountains can be explained by considering that buoyant updrafts over their ridgetops are on average relatively narrow. Thus, they are more strongly affected by the turbulent entrainment of environmental air, which depletes their moisture and cloud water content and makes them less effective at initiating deep convection. In our simulations, convective updrafts over moderately steep mountains, on the other hand, gain more moisture from the vapor flux at cloud base and lose less moisture due to horizontal vapor fluxes over the course of the day, leading to significantly higher moisture accumulation. The precipitation efficiency, a measure of how much of the condensed water eventually precipitates, is also considerably larger over the moderately steep mountains. The weaker convection over steep mountains is a robust finding, valid over a range of background environmental stratifications and mountain sizes.
Matthias Göbel et al.
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2023-648', Anonymous Referee #1, 05 May 2023
This manuscript uses idealized simulations to describe and quantify the sensitivity of orographic moist convection to terrain steepness. Experiments are performed with two different mountain heights and two different terrain slopes, allowing the two effects to be roughly separated. The simulations suggest that wider mountains generate thermal circulations that are more successful at accumulating moisture in the subcloud dry updraft, and that the moist convection over wider mountains more efficiently converts condensate into precipitation.
I found this manuscript to be well written and interesting, with precise analyses that help to gain insight into the problem. Therefore, I think it will provide a useful addition to the literature. However, I have some concerns about some of the analyses and associated interpretation, particularly regarding the cloud-layer budget analysis, the distinctions between subcloud and cloud-layer convection, and the resolvability of the orographic clouds. Therefore, I have comments that would be classified as “major”, which I hope the authors will address in their revision.
Major comments
- I found the cloud-water budget analysis of section 4.2 to be rather complicated and not as insightful as it could have been. I particularly wonder why the authors chose to make the lateral boundaries of the control volume vary in time. Normally this is avoided, and it adds a term to the budget with a comparable magnitude to the other large terms. Why not just fix the boundaries of the control volume based on the mountain dimensions, as done by Demko and Geerts (2010)? That would be much simpler, and it would help to expose the processes that matter the most without interference from the changing control volume.
- Beyond the above concern, there is limited insight into cloud-layer processes. The budget has value, but ultimately it is only used to show that (i) the less-steep mountains cause larger moisture fluxes into the cloud layer due to their wider updrafts, which follows directly from section 3.2 (which, by contrast, was more rigorous), and (ii) the clouds that form over the steeper mountains tend to evaporate rather than making much precipitation. From this result, the authors conclude that the narrowness of the clouds over the steeper mountains is what prevents them from intensifying. The clouds do look narrower in Fig. 3 but that result isn’t confirmed quantitatively. Also, the role of cloud width on cloud life cycles is not necessarily settled scientifically. It is a hypothesis that has been refuted in some studies (e.g., Dawe and Austin, ACP 2013). Therefore, I am not convinced that the authors’ conclusion is justified. Fortunately, there are many useful analyses that one can do to gain insight into cloud processes. I’ll start with the simplest: conditionally average (in the horizontal) a conserved variable (e.g., moist static energy or equivalent potential temperature) within the clouds and compare the profiles between different cases. This would show whether the clouds over steeper mountains (i) originate with lower moist static energy (or theta_e), and how they evolve with height. My guess is that these variables are already smaller at cloud base, and decrease faster with height, over steeper mountains, which would imply reduced buoyancy potential and faster dilution with height.
- The orographic clouds over the steeper mountains are so narrow in Fig. 3 that they may be under-resolved. Looking closely at the figure, the cloud appears to be < 500 m wide. With a horizontal grid spacing of 100 m, this cloud would be poorly resolved even in an inviscid model. But WRF is highly diffusive, with an effective resolution of 5-10dx, which means the cloud is poorly resolved. Thus, the cloud is likely subject to strong enough numerical diffusion to affect its development. If the authors have the resources, I suggest redoing the steeper mountain cases at 50 m or run a third case with an even wider mountain at 100 m. The latter would allow you to compare two cases (s10 and, say, s5) where the slope differs but the clouds are well resolved in both cases. This would also be useful for showing the robustness of the experimental trends, beyond the two cases (for a given peak height) currently conducted.
Minor comments
- L. 76: I’m not sure if the “local thermodynamic profile” is the most appropriate here. I would suggest replacing it with the “environmental thermodynamic profile”, because the orography locally changes the profile. If we’re thinking of a simple cause and effect relation, it is better to separate the environmental flow from the locally modified orographic flow.
- L. 117-118: I find the motivation of this study to be rather vague and not very exciting: “we continue investigating the impact of terrain geometry on orographic CI and take a closer look at strongly inhibited conditions”. This objective says what the authors intend to do, not the scientific problem they intend to address. It is better to frame an objective around a scientific question, because the methodology becomes a step toward achieving that goal. The methodology should not be the end unto itself.
- L. 135-136: I find it a bit odd that you would use positive-definite advection for scalars and WENO for vectors. Normally one would use WENO on scalar advection to avoid spurious oscillations that degrade the simulation of clouds. Can you please justify these choices? I’m guessing you used positive definite advection to ensure moisture conservation, which makes sense, but the monotonic and WENO options also do quite well at conserving moisture (I believe). If you see spurious oscillations forming near cloud edges in your simulations, it is likely owing to the PD scheme used for scalars.
- L. 231-232: the authors discuss the relative humidity diurnal evolution, but they don’t seem to show it. When discussing quantities are not shown, the reader should be informed of this via “(not shown)” or similar.
- Figures 7-8: The authors show these nice budgets but don’t clarify where the budgets are measured, which is a pity. The authors only say that they are “above the ridge”, which is rather vague to me. Do you mean the grid point over the mountain crest? Please specify that, because it's critical context.
- L. 267: “final” -> total
- L. 281-288: Are you confident that Fig. 9 allows you to calculate updraft width? It concerns me that the authors are measuring mass flux over the entire column and using that evaluate the width of the surface-based updraft. What if there is compensating mass flux aloft and/or cumulus convection? Both of these are not directly related to the surface updraft width but would figure into the column-integrated mass flux. On L. 288, the authors state that cloud latent-heat release leads to a widening of the updrafts. I don’t think that conclusion is justified, because the authors make no attempt to distinguish dry and moist updrafts in this calculation. It could be that the dry updrafts don’t change but moist convection aloft changes the mass flux profiles. This is an example of the authors reaching conclusions without acknowledging the limits of their analyses (as in major comment #2).
- L. 358: To me at least, the term Epbl in the moisture budget seemed irrelevant at first, since it relates to evaporation outside the control volume. I was only able to understand its value by realizing that the authors are probably only using it to infer, based only on measurements of *surface* precipitation (P) and microphysical tendencies, how much precipitation sediments out of the control volume, which is P+Epbl. The authors do not discuss this point, which I think invites confusion. As I mentioned in major comment 1, I think the complexity of the deformable control volume is a shortcoming, and this confusion compounds that effect.
- L. 378-379: Related to major comment 2, the authors seem to draw this conclusion out of thin air. They claim that the moist updrafts over the less-steep mountains are wider than those over steeper mountains, which hasn’t been quantitatively shown. They then conclude that these updrafts are more susceptible to entrainment, which they also haven’t shown...but I agree is indirectly implied by the moisture budget. Could it just be that these updrafts, due to lower moisture content at cloud base, are unable to generate as much adiabatic buoyancy, and are thus more susceptible to any suppressive effect (entrainment, detrainment, adverse pressure perturbations, etc.)? Decreased initial updraft moisture over steeper isladns was suggested by the dry moisture budget of the subcloud layer. This just seems like a conclusion that is stated much more strongly than is justified by the data, and alternative hypotheses are not considered.
- L. 422: The authors must appreciate that moist and dry updrafts behave differently and probably should not be lumped together. This sentence refers to narrower updrafts losing “moisture and cloud droplets”, not specifying whether the updrafts are dry or saturated. The presence of cloud droplets assumes saturated updrafts, but “thermal plumes” earlier in the sentence implies dry updrafts. Which one is it, or is this statement general enough for both? This problem reappears on L. 438, where the term “convective updrafts” is used without specifying dry or saturated.
- L. 443-444: I agree with this conclusion, but I think you could elaborate on it a bit. The fact that you have shown that you can reproduce Imamovic et al’s (2019) volume scaling at lower CIN gives you much more credibility in declaring the conditions under which it doesn’t work. I think that point is worth highlighting because it strengthens your conclusion.
- L. 449: A pet peeve of mine is the description of orography as “3D”. How can orography be 3D when it only varies at most in two directions (x and y)? A function of x and y is 2D, not 3D.
References:
Dawe, J. T., and P. H. Austin, 2013: Direct entrainment and detrainment rate distributions of individ- ual shallow cumulus clouds in an les. Atmos. Chem. Phys., 13 (15), 7795–7811, https://doi.org/ 10.5194/acp-13-7795-2013.
Citation: https://doi.org/10.5194/egusphere-2023-648-RC1 -
RC2: 'Comment on egusphere-2023-648', Anonymous Referee #2, 15 May 2023
The authors present a detailed look at how convection initiation changes with changes to terrain characteristics. The manuscript presents novel concepts that fit well within a growing body of work investigating convection initiation. To the best of my knowledge, no major flaws exist in the author reasoning or experimental setup. Below are several suggestions that I think can help strengthen the manuscript, but overall the manuscript is well written and the work appears to be high quality.
Principal criteria
Excellent (1)
Good (2)
Fair (3)
Poor (4)
Scientific significance:
Does the manuscript represent a substantial contribution to scientific progress within the scope of Weather and Climate Dynamics (substantial new concepts, ideas, methods, or data)?X
Scientific quality:
Are the scientific approach and applied methods valid? Are the results discussed in an appropriate and balanced way (consideration of related work, including appropriate references)?X
Presentation quality:
Are the scientific results and conclusions presented in a clear, concise, and well-structured way (number and quality of figures/tables, appropriate use of English language)?X
I did not identify any major concerns that would warrant extra work from the authors. The paper is well written, contains well-reasoned scientific design and analysis, and reaches conclusions that are supported by the data presented. Below some moderate and minor concerns are mentioned. Moderate concerns are ones which I think might require a little additional work or explanation, but I do not believe are major hinderances to publication. Minor concerns relate to grammar or figure clarity and can be addressed by the authors as they see fit.
Moderate concerns:
- The primary conclusion is that wider terrain leads to stronger convection. This is tied to the initial updrafts being wider and less resistant to entrainment. It was not immediately clear in reading the manuscript why the wider terrain leads to wider updrafts. Perhaps this is common knowledge in mountain meteorology and I am not aware of the reason, but it would be nice to make the connection obvious in the paper. If there isn’t a well-established reason for this, it could be helpful to speculate as to why.
- The horizontal grid spacing for the main simulation suite that the manuscript is based on is largely in line with that of other recent manuscripts (e.g., Morrison et al. 2021, Peters et al. 2020). While this might be on the marginal side for resolving the influence of entrainment on smaller clouds/thermals like the ones depicted here, I have little reason to believe that the results of the study would change if a finer resolution were to be used. The runs at 250 m grid spacing contain clouds that are marginally resolved; however, the conclusions drawn from these simulations also fit within the conceptual framework that the manuscript develops, and I also think that the conclusions drawn from these sensitivity tests would not change at an increased spatial resolution.
- The authors may wish to discuss some of their results in the context of recent publications investigating updraft width and entrainment and their eventual influence on deep convection initiation including Morrison et al. (2021) and Peters et al. (2020).
Minor comments
Line 7: I found the use of relatively steep and moderately steep confusing throughout the manuscript and had to keep reminding myself of which was which. Perhaps more clear language could be used like steeper and less steep? I am open to leaving it as is, or leaning more on the s10 and s20 references, but did want to note that it was confusing to me.
Line 225: Would calculating these convective indices with mixed layer parcel properties (generally calculated over the lowest 100 mb) be more predictive? Sometimes these are used to account for entrainment, although they aren’t always perfect. I am not very familiar with this parameter, but given the focus on entrainment, using entraining convective available potential energy (ECAPE, https://arxiv.org/abs/2301.04712) may be predictive here. I’m not sure how useful it is though, given there us very little storm relative flow here. If you do not think this would be applicable or useful, I am fine ignoring this suggestion.
Citation: https://doi.org/10.5194/egusphere-2023-648-RC2
Matthias Göbel et al.
Data sets
Data and code for "Adverse impact of terrain steepness on thermally-driven initiation of orographic convection" Matthias Göbel https://doi.org/10.5281/zenodo.7794166
Matthias Göbel et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 148 | 60 | 5 | 213 | 1 | 0 |
- HTML: 148
- PDF: 60
- XML: 5
- Total: 213
- BibTeX: 1
- EndNote: 0
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1