the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 17 Sep 2025
Impact of small-scale orography on deep boundary layer evolution and structure over the Tibetan Plateau
Abstract. We investigate how small-scale orography influences the evolution and structure of the exceptionally deep convective boundary layer (CBL) over the Tibetan Plateau (TiP). Using large-eddy simulations (LES) at 50 m resolution under semi-idealized dry conditions, we compare three experiments over an elevated plateau (4.2 km above sea level, ASL): FLAT (no orography), REAL (realistic terrain), and REALu10 (REAL plus an upper-level wind of 10 m s⁻¹). All simulations produce very deep CBLs, reaching ~9 km ASL by late afternoon, consistent with record-high values observed over the TiP. Small-scale orography substantially accelerates early growth: by midday the CBL in REAL is ~80 m higher than in FLAT, and locally above the mountain it is ~500 m deeper. This terrain-induced advantage narrows later in the day as all cases attain extreme depths. In REALu10, the CBL grows even higher, reaching 9.4 km ASL by 19 LT. Added shear organizes convection into longitudinal roll vortices, contrasting with the isolated terrain-anchored plumes in the no-shear case, and enhances vertical mixing.
These results demonstrate that unresolved small-scale orography can increase daytime CBL height by up to 15 % and expedite entrainment of free-tropospheric air. Under clear-sky conditions, the plateau’s CBL can exceed 9 km within a single day given strong surface heating and weak stability aloft. Our findings highlight the importance of including fine-scale terrain and shear effects in models, as their omission may underestimate CBL growth and vertical exchange over high-altitude regions.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(17854 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(17854 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-4302', Anonymous Referee #1, 05 Oct 2025
-
AC1: 'Reply on RC1', Ivan Bašić, 10 Dec 2025
Thank you very much for reading the manuscript thoroughly and for your excellent (and very clear) insight. We hope this response is satisfactory. Further comments and/or suggestions regarding the revised manuscript are welcome.
===============
Major comments
===============- The imposed wind profile is used only in the sensitivity experiment REALu10 (and now in FLATu10 as well) and is not intended to represent a realistic valley–ridge wind structure. Our main motivation was to introduce a non-zero, large-scale background wind within the boundary layer, with a magnitude comparable to the conditions documented by Chen et al. (2016, their Fig. 5). We set the initial wind to zero near the surface to avoid exciting strong orographic waves during model initialization. As the simulation evolves, there is downward momentum transport resulting in a non-zero mean wind within the BL and a shear layer across the entrainment zone.
We revised section 2.2 to briefly explain this; and section 3.1 where comment this evolution (Fig. 2) in more detail. - In our configuration, the primary ingredient enabling extreme CBL growth is the weak stability aloft, associated with the passage of an upper-level trough (PU anomaly) (Chen et al. 2016). Further ingredients include the high elevation (lower air density), the arid land surface, and mechanical forcing (in REAL and REALu10) that together favor strong dry convection. We agree that the same mechanism—weak stability combined with strong surface heating and small-scale orography—can also operate over lower plateaus or desert regions, even though the resulting CBLs would not reach the extreme heights found on the TiP. Thus, our FLAT vs REAL comparison should be understood as illustrating a general mechanism whose qualitative behavior is transferable to lower plateaus, while the quantitative CBL depths will depend on elevation, background stability, and surface forcing. We have clarified this point in the revised conclusions that a systematic exploration for other plateau regimes remains a topic for future work.
===============
Minor comments
===============- l.16 TiP abbreviation was revised as suggested. We also applied it to the other abbreviations, like CBL.
- l.70 "FLAT (no orography)" -> "FLAT (no local orography)"
- l.91 revised as suggested
- l94 removed repetition
- l.96 semi-slip: added a line in section 2.1
- l.98 idealized sounding: revised as suggested in section 2.1
- l.103 revised: in section 2.1 we write the simulation start and how long it runs.
- Table 1: revised as suggested.
- l.139 to quantify -> to determine
- Fig. 3:
-"2.32.3.2" -> "2.3.2". Added in caption in Fig. 3: "(...), with the 800 m APL orography isoline indicated by a black contour" (APL = above plain level)
- we write only 6 km ASL - Fig. 4 "at 15 LT" added.
- Eqs (4) and (5): revised u'^2 and v'^2
Added resolved variance part (eq 4) in 2.3.1. We refer to this equation in 3.3.
We have majorly rewritten this section so it is hopefully clearer now. - Fig. 7:
- we revised this so there are no tracer 1 or 2, but surface-based and upper-level sect 2.1.
- the units of the isolines are specified now, and we changed the units of the upper-level tracer to dimensionless percentage. - Fig. 8 :
- Caption is revised.
A quick explanation (just in case, since the previous caption was wrong and misleading):
1. we ensemble average the surface-based tracer at 15 and 19 LT. (Fig. 7 shows how it looks).
2. we diagnose the ABL heights from this (500*500 columns == 500*500 ABL heights).
3. we perform statistics on the ABL heights. - l.249 revised this section so it uses percentages instead of meters, and uses meter resolution (and percentages) in table 2. We introduced different height abbreviations:
APL (above plain height), MSL (above mean sea level), AGL (above ground level, i.e., "above terrain height"). The mentioned percentages refer to ABL height APL. - appendix 1: we rewrote section 2.3.2 to mention the parcel method and refer to the appendix for a detailed/quantitative comparison between the methods. We chose to focus on the tracer method and keep the parcel method in the appendix.
Citation: https://doi.org/10.5194/egusphere-2025-4302-AC1 - The imposed wind profile is used only in the sensitivity experiment REALu10 (and now in FLATu10 as well) and is not intended to represent a realistic valley–ridge wind structure. Our main motivation was to introduce a non-zero, large-scale background wind within the boundary layer, with a magnitude comparable to the conditions documented by Chen et al. (2016, their Fig. 5). We set the initial wind to zero near the surface to avoid exciting strong orographic waves during model initialization. As the simulation evolves, there is downward momentum transport resulting in a non-zero mean wind within the BL and a shear layer across the entrainment zone.
-
AC1: 'Reply on RC1', Ivan Bašić, 10 Dec 2025
-
RC2: 'Comment on egusphere-2025-4302', Anonymous Referee #2, 07 Oct 2025
The authors conduct semi-idealized LES experiments under dry conditions to compare three scenarios: a flat plateau (FLAT), realistic terrain (REAL), and realistic terrain with added upper-level wind (REALu10). They analyze boundary layer depth, turbulence structure, tracer dispersion, and vertical mixing. Key findings include:
Small-scale orography accelerates early CBL growth and leads to localized deepening (up to ~500 m over mountains). Wind shear organizes convection into roll vortices, enhancing mixing and further increasing CBL depth (up to 9.4 km ASL). Terrain and shear effects are roughly additive, with the deepest and most uniform CBL occurring in REALu10. A tracer-based method for diagnosing CBL height is introduced and validated against a parcel method.
The manuscript in its current form has minor weaknesses that must be addressed before it can be considered for publication. These primarily concern a lack of methodological clarity, and occasional overstatement of conclusions.
This study makes an original contribution by quantifying the role of small-scale orography and shear in driving extreme CBL depths over the Tibetan Plateau. I am curious, compared with the stability aloft, the fine-scale terrain and shear effects are still the firs-order factor for the extreme high CBL? Please add more sensitivity simulation to compare their respective importance.
The description of the Terrain Processing methodology is vague and omits critical details necessary for reproducibility and full assessment. The statement that the terrain was "tapered" and then "smoothed with a Gaussian filter (200 m)" is insufficient. What was the standard deviation of the Gaussian kernel? Was the filter applied once or iteratively? A figure showing the original SRTM data, the tapered terrain, and the final smoothed terrain used in the REAL case would be immensely helpful to understand the degree of modification and its potential impact.
Flow Decomposition: The description of the Reynolds averaging procedure is confusing. The authors use a 2D Gaussian filter on time-averaged fields. This is a spatial filter, not a classic ensemble average. The justification for this hybrid approach over a pure temporal or spatial average needs to be strengthened. Furthermore, the choice of a 500 m standard deviation and a 4 km support width seems arbitrary; Please give an explanation why these values were used. A sensitivity analysis or a stronger justification based on the dominant turbulent length scales would be beneficial.
Initial Conditions and Forcing: The use of an idealized sounding with a constant weak lapse rate above a 300 m mixed layer is a major simplification. The authors must more explicitly discuss the limitations this imposes, as the real atmosphere over the TP likely has multiple layers of varying stability that could significantly modulate entrainment and CBL growth.
Local vs. Domain-Averaged Impacts: While the local enhancement over the ridge (R-TOPO) is well-quantified (~500 m), the domain-averaged impact of orography (REAL vs. FLAT) is surprisingly small (~80 m). The authors should discuss this dichotomy more thoroughly. Does this imply that the overall impact of small-scale orography on the plateau-scale CBL is minor, but it creates strong local heterogeneity? The current narrative emphasizes the importance of orography, but the domain-mean results tell a more nuanced story.
The REAL terrain is described as featuring slopes up to 30° and ridges up to 1400 m. It would be helpful to contextualize this within the broader Tibetan Plateau to justify its representativeness.
While the idealized setup is appropriate for isolating mechanisms, the authors should more explicitly discuss how the exclusion of moisture, clouds, and synoptic variability might affect the generalizability of the results.
The authors mention that unresolved orography may bias CBL representation in models. A more quantitative discussion—e.g., how much CBL growth is underestimated in models that smooth terrain—would enhance the impact of the study.
The results section frequently references figures (e.g., Fig. 3, 4, 7) before the reader has any context for what these figures show. The narrative would be much clearer if the description of each figure was immediately followed by the interpretation of its results.
The abstract and conclusions are somewhat repetitive. The conclusions could be strengthened by synthesizing the findings into a broader conceptual model of how terrain and shear interact, rather than just restating the results.
Line 6, Small-scale orography substantially accelerates early growth: by midday the CBL in REAL is ∼80 m higher than in FLAT, 80m higher than FLAT is substantially? Compare surface flux and stability in the free air, is it still substantially?
Line 7, locally above the mountain it is ∼500 m deeper. This sentence is confusing, please rephrase it.
Line 7, This terrain-induced advantage narrows later in the day, what is narrows? Please rephrase this sentence.
Line 12, Under clear-sky conditions, the plateau’s CBL can exceed 9 km within a single day given strong surface heating and weak stability aloft. This conclusion has been approved by Chen` study. It may not be proper to be included in the abstract and taken as a conclusion of this study.
Line 18, One distinguishing feature of the TiP’s ABL is its extraordinary
depth. It`s better to cite the reference here: X., C. and Y., M., 2025. The unique atmospheric boundary layer over the Tibetan Plateau (in Chinese). Chin Sci Bull, 70: 4180–4187.
Line 37, it is better to further point out the weak stability in the lower troposphere is associated to the tropopause folds activity (Chen et al. 2011 ACP).
Chen, X., Ma, Y., Kelder, H., Su, Z. and Yang, K., 2011. On the behaviour of the tropopause folding events over the Tibetan Plateau. Atmos. Chem. Phys, 11: 5113-5122.
P3, L75: Specify what "REAL terrain" represents in terms of maximum height AGL and horizontal scales.
P4, L105: The vertical grid spacing is described twice with slightly different wording. Consolidate this into a single, clear description.
P9, eq. 4-6, The equations for TKE components are incorrectly formatted. w'^2 is repeated for all three components. This must be corrected to u'^2, v'^2, and w'^2.
Citation: https://doi.org/10.5194/egusphere-2025-4302-RC2 -
AC3: 'Reply on RC2', Ivan Bašić, 10 Dec 2025
We thank Reviewer 2 for their constructive and thoughtful comments. We appreciate the reviewer’s recognition of the manuscript’s strengths and their helpful suggestions for improving clarity and moderation of our conclusions. We have carefully addressed all points and revised the text accordingly.
==========
Responses
==========- The weak stability aloft is indeed the primary control on extreme CBL growth, as it sets the background conditions for deep dry convection (see energy balance model in Chen et al.). The present study focuses on how small-scale orography and wind shear further modulate this growth under otherwise identical, weak-stability conditions. In other words: small-scale orography and shear are the sensitivity test to the already favorable conditions for deep CBL growth.
In the manuscript, we mention this in section 2.1 and make it clearer overall. - We revised the terrain processing methodology (section 2.2), added Appendix B discussing super-gaussian filter, and added additional panel in Fig. 1 for more clarity.
- The 2D Gaussian filter.
It was applied to isolate sub-mesoscale variability in the time-averaged fields, serving as a spatial decomposition rather than a strict Reynolds average. The so-called "ensemble average" is then used. The filter parameters (σ = 500 m and support width = 4 km) were chosen deliberately to match the dominant turbulent length scales observed in the simulations, corresponding to the typical horizontal size of large thermals and roll vortices in the CBL.
Additional tests with smaller (σ = 250 m) and larger (σ = 1 km) smoothing lengths yielded consistent qualitative patterns, confirming that the decomposition is not sensitive to the exact parameter choice. We have clarified this rationale in the revised manuscript.
Why did we chose to include spatial-average as well as the time-average? With only the time averaging, the turbulent statistics are poor - but domain-means are not applicable due to orography. So we are choosing time averaging AND 2D Gaussian, to improve turbulent statistics. This approach is a "poor-man's" approximation to ensemble-average. - Initial conditions and forcing.
The weak lapse rate above the 300 m mixed layer was indeed chosen intentionally to reproduce the favorable thermodynamic conditions observed during deep boundary-layer development over the Tibetan Plateau. In particular, Chen et al. (2016) report weak free-tropospheric stability (near-dry-adiabatic lapse rates) during strong surface heating episodes, which we use as a representative reference.
We acknowledge that real soundings often exhibit multiple stable layers and inversions; however, simplified profile was chosen to isolate the effects of local orography and shear without introducing additional complexity from vertical layering. The goal was to conduct a sensitivity study, not reproduce observations accurately. We have clarified these limitations in the revised manuscript. - local vs. domain-averaged impacts.
You are exactly right and we tried to point it out in the manuscript, but it was not emphasized enough. What you observed reflects the fact that the orographic influence is spatially heterogeneous — it strongly amplifies local entrainment and turbulence over elevated slopes but averages out when integrated over regions that also include valleys and flat areas.
We have clarified in the revised results section and conclusions that subgrid orography mainly enhances spatial variability rather than substantially altering the domain-mean CBL height.
Also, we added a new Fig. 9 where we show how the boundary layer is actually less deep over orography - although may reach higher heights. This distinction is made clear (height above ground level vs. heigh above mean sea level). - terrain representativeness.
The chosen terrain represents a typical ridge–valley structure within the central Tibetan Plateau, with elevation differences of 1000–1500 m and slopes up to ~30°, consistent with regional SRTM data. The periodic boundary conditions ensure that the domain effectively repeats this representative mountain structure, rather than emphasizing any single extreme feature. The local context of the selected region is already illustrated in Fig. 1a, where the surrounding orography can be seen. We have added this clarification to the Methods section.
In the revised manuscript, the slopes go up to ~36°. - idealized setup.
We have expanded the discussion of limitations in the Conclusions to more explicitly address how the exclusion of moisture, clouds, and synoptic variability affects the generalizability of our results. As suggested, we now clarify that the objective of our idealized setup was to isolate the mechanisms by which orography and shear influence CBL development, and that these dynamical mechanisms can reasonably be generalized to other high-elevation or weakly stratified environments. We also note that moist convection and cloud processes would likely enhance vertical exchange and turbulence compared to our dry configuration, potentially amplifying some of the effects identified here. Finally, we specify that the synoptic background in our case corresponds to unusually weak stability, which is a key condition favoring deep CBL growth, and we point to future work needed to explore a wider range of synoptic regimes. - unresolved orography.
We have added a short quantitative statement in the discussion to emphasize this implication. Also, in section 3.4. we elaborate on this and clarify that models underestimate extreme BLs, and that relative ABL depth is smaller over the mountains (new Fig. 9). - figure referencing.
Thank you for pointing this out. We moved the figures so they appear earlier in the document, i.e., before they are referenced. - abstract and conclusions repetitive
Abstract is revised.
In conclusions: added future guidelines and also how can it be expanded. - lines 6 and 7.
We rephrased this part in the abstract to (also) include some ideas from the previous comment about local vs. domain-averaged impacts. - line 12.
We revised the abstract. - line 18.
Reference added. Thank you for the suggestion. - line 37.
The reference has been substituted. - line 75, 105, and equations 4-6.
All revised as suggested.
Citation: https://doi.org/10.5194/egusphere-2025-4302-AC3 - The weak stability aloft is indeed the primary control on extreme CBL growth, as it sets the background conditions for deep dry convection (see energy balance model in Chen et al.). The present study focuses on how small-scale orography and wind shear further modulate this growth under otherwise identical, weak-stability conditions. In other words: small-scale orography and shear are the sensitivity test to the already favorable conditions for deep CBL growth.
-
AC3: 'Reply on RC2', Ivan Bašić, 10 Dec 2025
-
RC3: 'Comment on egusphere-2025-4302', Anonymous Referee #3, 07 Oct 2025
This manuscript investigates the impact that small-scale orography features have on the evolution and structure of the convective boundary layer over the Tibetan Plateau. The author used an LES model (CM1) and ran three idealized simulations for a special case using three different configurations that differ in the small-scale orography features included and in the presence/absence of upper-level wind. The results show these small-scale features and shear effects enhance the CBL growth by up to 15%. The paper is well designed, clearly states the objectives and methodology, and the results are properly presented and discussed. I recommend the paper for publication, although some minor comments should be addressed before that.
Minor corrections
- Line 4: replace “a upper-level” by “an upper-level”.
- Line 8: please explain what LT means (first time that appears in the text).
- Line 16: please explain what the “TiB” abbreviation means.
- Line 19: same for the abbreviation “CBL”.
- Line 94 to 95: sentence repeated (already explained in line 91).
- Line 96: please include a brief explanation of what semi-slip condition means.
- Line 116: I think “FLAT” is a bit confusing cause this simulation is not flat itself, but does not include small-scale features. Please consider changing the name, like NOSSO or something that states more clearly how the orography is represented in that simulation.
- Section “Experiments and Data”: could the authors include whether the simulations were run with a spinup period?
- Lines 145-147: these two sentences say the same in a slightly different way. Please consider keeping just one to avoid repetition.
- Line 153: include the symbol q after potential temperature, as this symbol will be used in the caption of several figures throughout the manuscript.
- Figure 2: this figure is just briefly discussed in lines 154 and 155. I would suggest moving it to the Appendix.
- Line 157: add “w” after vertical velocity as this will be used in the caption of the following figures.
- Figure 3: in the caption, the authors mentioned that the vertical cross-sections are calculated for y=0. To improve the clarity of the manuscript, I suggest including a line in figure 1 showing where this cross-section is located.
In the caption of this figure, there is a reference to 2.32.3.2. Please correct. - Line 182: the authors say they also analysed the instantaneous vertical velocity at lower levels and that those results are not shown. Perhaps the one at 500 m AGL could be included in the Appendix.
- Line 189 and 190: w’2 should be replaced by u’2 and v’2, respectively.
- Figure 7: please indicate the units used and explain the meaning of the lines shown.
- Line 202: FLAT should be replaced by R-FLAT if I am not mistaken.
- Figure 8: if I am not mistaken, y-axis is showing the altitude of the CBL height and not the surface-based tracer concentration. Please clarify this.
- Line 281: the Appendix is never mentioned throughout the manuscript. Perhaps, you could refer to it in section 2.3.2 when talking about the method used to diagnose the boundary layer height.
Citation: https://doi.org/10.5194/egusphere-2025-4302-RC3 -
AC4: 'Reply on RC3', Ivan Bašić, 10 Dec 2025
We thank Reviewer 3 for the comments and corrections. The attention to detail is really helpful and it improved the manuscript.
==========
Responses
==========- line 4,... line 96: revised as suggested.
- line 116:
When introducing FLAT, we write: FLAT (no local orography). - spinup period.
The first 2 hours are considered spinup. They are discarded as the model spinup to achieve a reasonable physical atmospheric state. We added a sentence to section 2.1 regarding this. - lines 145-147: revised.
- line 153: revised
- Figure 2.
The authors consider Figure 2 to be an important part of the paper clarity and reading flow and we decide to keep it as-is - but we expanded the discussion around it in Section 3.1. - line 157: revised
- Figure 3.
The authors think adding the horizontal line is unnecessary, and that the one added in the top row of Fig. 3 serves this purpose. However, if the reviewer insists, we would add the horizontal line.
Caption is revised. - line 182.
We added Fig. C1 (in the appendix) with four panels showing two heights AGL: 250 m & 500 m. - line 189-190: revised
- Figure 7.
We added the units and we revised the figure so that the upper-level tracer is dimensionless (percentages) rather than actual values, since it is more helpful for a qualitative read. - line 202: the authors carefully checked the line, and "FLAT" is correctly placed.
- figure 8: revised.
- line 281: revised as suggested.
Citation: https://doi.org/10.5194/egusphere-2025-4302-AC4
-
RC4: 'well-written manuscript, but more analysis on TKE budget neeed', Anonymous Referee #4, 08 Oct 2025
Dear authors
As a first note, I want to inform you that I already read through the three other review statements. In general, I agree with all of them, and henceforth, my comments might be less than a typical review, but please consider them as equally relevant.
General comments
This is a clear, well-written manuscript exploring the reasons for a very high-ranging boundary layer over the Tibetan Plateau. Given the Tibetan Plateau's role in the global climate system, this research is relevant for the community. The authors employ the CM1 model for idealized simulations and analyze the boundary-layer structure in simulations with/without orography, and imposing different background wind conditions. The manuscript is well-written, but additional analysis (TKE budget) and possibly one more simulation (FLAT/u10m) are necessary to strengthen the conclusions of the authors and to make the MS suitable for publication in ACP.
Major comments
You identified shear-induced turbulence as one of the major reasons for the high-ranging ABL over your plateau. However, although you mention shear throughout the manuscript, you never show its actual quantity, e.g., by calculating the shear production term in the TKE budget equation compared to buoyancy forcing. Creating a figure and the accompanying text of the TKE budget will strengthen your manuscript massively.As a second point, you talk about 'terrain forcing'. I understand that this conclusion is drawn from a comparison of flat vs. complex terrain. However, if I understood it right, the FLAT simulation is a sub-domain of the domain with real toporgaphy.
1) How can you make sure that the FLAT subdomain is not affected by a plain to mountain flow towards your plateau?
2) As a second point, it would be also advisable to add a simulation/analysis of a flat-terrain, but shear-induced simulation, i.e., FLATu10, so that you can make a direct comparison with the plateau simulation - and - especially relevant - to isolate the 'terrain' and 'shear' effects accodingly.Minor comments
As other referees already pointed out minor formulation or speeling errors, I would only want to encourage the authors to avoid statements like 'typical development' or 'expected behaviour' without citing the relevant literature. Not all readers are aware of how boundary-layer development over orography is 'supposed to look like'. Furthermore, the last paragraph of the introduction might better fit in the methods Section.Therefore, I recommend major revisions for the manuscript and look forward to the revised version.
Best regards
Citation: https://doi.org/10.5194/egusphere-2025-4302-RC4 -
AC5: 'Reply on RC4', Ivan Bašić, 10 Dec 2025
Thank you for your valuable comments and insights, and very helpful suggestions. We address each of them carefully.
===============
Major comments
===============- Shear production.
We add additional panel to Fig. 5, which was previously just a resolved TKE profile at 15 LT. Now it includes the shear and buoyancy production term at 15 LT as well. We expand the accompanying text in section 3.3. - Terrain forcing.
The FLAT simulation is a stand-alone simulation with no real topography. There is a subdomain called R-FLAT and it is only analysed in Fig. 8.
Answering "1)": there is no plain-to-mountain flow (see Fig. 1d). However, there is plain-to-mountain flow in R-FLAT, but it is intentional since it is a subdomain and analyzed as such.
Answering "2)": we added an additional simulation to the study: FLATu10 to do exactly what the reviewer suggested, and it helped us show our concluding points more clearly.
===============
Minor comments
===============- The authors added the relevant literature (Stull, 1988) and rephrased the sentence that mentions the typical TKE profile in the late afternoon to make the statement more clear.
- Last paragraph of the introduction was rewritten to only contain an overview of what was done with no details. The details were moved to the Methods section, like the reviewer suggested.
Citation: https://doi.org/10.5194/egusphere-2025-4302-AC5 - Shear production.
-
AC5: 'Reply on RC4', Ivan Bašić, 10 Dec 2025
-
AC2: 'General note to the referees', Ivan Bašić, 10 Dec 2025
Dear referees,The manuscript has undergone major revisions, particularly in the Methods and Results sections. The main changes are:
- Addition of FLATu10: A new experiment (FLATu10) was added to better isolate the impacts of orography and wind shear.
- Background wind in FLAT and REAL: A weak background wind was introduced in FLAT and REAL (constant profile of 1 m/s) to break the W–E symmetry. This modification was successful (see Figs. 3, 6, and 7).
- Terrain smoothing settings: The standard deviation σ of the 2D Gaussian filter applied to the terrain was reduced from 200 m to 75 m (a factor of 4 to 1.5 grid spacings) to retain more small-scale terrain features. While the overall results do not change, the differences are noticeable in 99th percentiles, which reinforces our main conclusions.
- New Appendix B and additional figure: Appendix B was added to describe the super-Gaussian tapering of the lateral boundaries, and Fig. C1 now shows w' perturbations at lower levels, which were previously only mentioned in the text.
- New ABL–terrain diagnostic (Fig. 9): Figure 9 was added to show ABL depth (km AGL) and ABL height (km MSL) as a function of underlying terrain elevation, providing additional insight into the influence of small-scale orography.
- Greater emphasis on the upper tail of the ABL height distribution: we highlighted the behaviour of the upper tail (99th percentile) of the ABL height distribution (Fig. 8; Table 2; Section 3.4; Conclusions). This clarifies how small-scale orography disproportionately affects the deepest convective columns.
Kind regards,
The authorsCitation: https://doi.org/10.5194/egusphere-2025-4302-AC2
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-4302', Anonymous Referee #1, 05 Oct 2025
Review in connection with the manuscript entitled
Impact of small-scale orography on deep boundary layer evolution and structure over the Tibetan Plateau
by Basic et al.
Manuscript: https://doi.org/10.5194/egusphere-2025-4302
General
In this paper the authors use high-resolution LES (CM1) in idealized experiments to study the impact of ‘subgrid-scale orography’ (subgrid to typical global NWP or regional climate models) on the development of extremely deep convective boundary layers on the Tibetan Plateau. It is found that beyond the earlier identified conditions of weak stratification aloft and not too weak surface heating, ‘local orography’ indeed has an enhancing effect on CBL growth – thus leading to order 10% higher CBL. Overall, the paper is well designed (numerical experiment) and the results are presented in a consistent manner. I have, however, two ‘major comments’ addressing some details of the experiment. They are called ‘major’ because they cannot be attributed to a line (as the minor comments) – but will easily be addressed. In this sense, I think , the paper can be published after having addressed the minor comments.
Major comment
- Why is mechanical forcing introduced in such an unrealistic way (i.e., the steep gradient from zero wind speed in the valley to 10 m/s over just 500 m)? If there are conditions of forced convection, i.e. nonzero mechanical turbulence, this will result in a more gradual decrease of mean wind speed towards the mean plateau level. Can the authors comment on that?
- The numerical experiment is placed over the TiP and the local orography is used to define ‘REAL’ case. However, the experiment has no other ingredient of an elevated plateau (with the exception of pressure, perhaps). The question is therefore, to which degree the conclusions can be transferred to lower plateaus. Clearly, an enhanced CBL height over a lower plateau would not lead to those record-high CBLs with the potential to even reach the stratosphere – but still, a several-km deep CBL has been observed either under conditions of strong convection at desert locations, or on the TiP. This aspect should be worked out to some detail in a revised version.
Minor comments
l.16 ‘TiP’: the abbreviation is introduced in the abstract – but this usually not counted. So, please introduce the abbreviation here as well.
l.70 ‘FLAT (no orography)’: this may be a matter of definition, but I consider the TiP to be part of ‘orography’, too. So, maybe the authors want to consider to re-label this to locally flat (‘no local orography’, no ‘subgrid-scale orography ‘) or alike.
l.91 this ‘delta20 m’ notation seems to be odd. Why not ‘the vertical grid spacing is set to 20 m below….’? Several occurrences on this and the next line.
l.94 ‘the vertical grid spacing…’: this is repeated (and this time the ‘delta’ is already removed….)
l.96 what is a semi-slip condition? Please explain.
l.98 ‘idealized sounding’: please specify whether this only refers to temperature, or also includes humidity and wind speed (the u10 simulation suggests that the reference has no wind?)
l.103 ‘…starting a 09LT….’: is this the start of the simulation? Or is there a spin-up period considered? Please specify.
Tab 1 ‘Flat orography’ seems to be a contradiction. In the text it is referred to as ‘no orography’, which also seems to be counter intuitive...(see comment to l.70). Again, I suggest to consider ‘no subgrid scale orography’, or ‘elevated plain’ or similar).
l.139 ‘to quantify’ is probably not appropriate here – consider ‘to diagnose’ or ‘to determine’ instead.
Fig.3, caption: what does ‘see 2.32.3.2‘ refer to? Also, the black contour line should be explained, as 800 m above plateau level (i.e., 4200 m ALS)
Fig.3 In the caption it says ‚at 1800 m AGL (6 km ASL)‘: for the ‚REAL‘ simulations the equivalence of these two statements is not given. Please specify, which applies.
Fig. 4, caption: must specify the time, when the instantaneous fields are obtained.
Eqs (4) and (5): the lhs must be u’^2, v’^2 respectively. Furthermore, if using the according to (1) and using (2) the ‘difference between the time-averaged second and first moments’ is not obvious. I suggest to explicitly derive this (maybe in an appendix). However, in the nomenclature of the present work, it would, correspond to the resolved variance. This should be added.
Fig.7 you have introduced the tracers as ‘1’ and ‘2’ – and here, they are referred to as ‘surface-based’ and ‘upper-level, respectively. I suggest to introduce this convention where the tracers are introduced. Also, in the caption of Fig. 7, the units of the isolines must be specified.
Fig.8, caption: ‘horizontal distribution...’: if I understand the accompanying text correctly, these are the horizontal statistics (median, interquartile range) of the diagnosed mixing height (or CBL height) based on the surface emitted tracer and the elevated tracer (and not: tracer concentrations). I furthermore cannot understand how ‘the mean concentration’ can be marked (with an x) in a height domain (time) diagram. Please explain.
l.249 ‘…by 383 m…’: I don’t think that ‘meter resolution’ here is appropriate (I much more like the ‘about 10%’, even if in the summary then it should be indicated percent of what). Putting into context could also be done by expressing the change in percent of the terrain height.
Appendix 1: this appendix is never mentioned in the manuscript, nor is the parcel method. So, either the appendix can be removed from the ms, or the parcel method is included into the analysis.
Citation: https://doi.org/10.5194/egusphere-2025-4302-RC1 -
AC1: 'Reply on RC1', Ivan Bašić, 10 Dec 2025
Thank you very much for reading the manuscript thoroughly and for your excellent (and very clear) insight. We hope this response is satisfactory. Further comments and/or suggestions regarding the revised manuscript are welcome.
===============
Major comments
===============- The imposed wind profile is used only in the sensitivity experiment REALu10 (and now in FLATu10 as well) and is not intended to represent a realistic valley–ridge wind structure. Our main motivation was to introduce a non-zero, large-scale background wind within the boundary layer, with a magnitude comparable to the conditions documented by Chen et al. (2016, their Fig. 5). We set the initial wind to zero near the surface to avoid exciting strong orographic waves during model initialization. As the simulation evolves, there is downward momentum transport resulting in a non-zero mean wind within the BL and a shear layer across the entrainment zone.
We revised section 2.2 to briefly explain this; and section 3.1 where comment this evolution (Fig. 2) in more detail. - In our configuration, the primary ingredient enabling extreme CBL growth is the weak stability aloft, associated with the passage of an upper-level trough (PU anomaly) (Chen et al. 2016). Further ingredients include the high elevation (lower air density), the arid land surface, and mechanical forcing (in REAL and REALu10) that together favor strong dry convection. We agree that the same mechanism—weak stability combined with strong surface heating and small-scale orography—can also operate over lower plateaus or desert regions, even though the resulting CBLs would not reach the extreme heights found on the TiP. Thus, our FLAT vs REAL comparison should be understood as illustrating a general mechanism whose qualitative behavior is transferable to lower plateaus, while the quantitative CBL depths will depend on elevation, background stability, and surface forcing. We have clarified this point in the revised conclusions that a systematic exploration for other plateau regimes remains a topic for future work.
===============
Minor comments
===============- l.16 TiP abbreviation was revised as suggested. We also applied it to the other abbreviations, like CBL.
- l.70 "FLAT (no orography)" -> "FLAT (no local orography)"
- l.91 revised as suggested
- l94 removed repetition
- l.96 semi-slip: added a line in section 2.1
- l.98 idealized sounding: revised as suggested in section 2.1
- l.103 revised: in section 2.1 we write the simulation start and how long it runs.
- Table 1: revised as suggested.
- l.139 to quantify -> to determine
- Fig. 3:
-"2.32.3.2" -> "2.3.2". Added in caption in Fig. 3: "(...), with the 800 m APL orography isoline indicated by a black contour" (APL = above plain level)
- we write only 6 km ASL - Fig. 4 "at 15 LT" added.
- Eqs (4) and (5): revised u'^2 and v'^2
Added resolved variance part (eq 4) in 2.3.1. We refer to this equation in 3.3.
We have majorly rewritten this section so it is hopefully clearer now. - Fig. 7:
- we revised this so there are no tracer 1 or 2, but surface-based and upper-level sect 2.1.
- the units of the isolines are specified now, and we changed the units of the upper-level tracer to dimensionless percentage. - Fig. 8 :
- Caption is revised.
A quick explanation (just in case, since the previous caption was wrong and misleading):
1. we ensemble average the surface-based tracer at 15 and 19 LT. (Fig. 7 shows how it looks).
2. we diagnose the ABL heights from this (500*500 columns == 500*500 ABL heights).
3. we perform statistics on the ABL heights. - l.249 revised this section so it uses percentages instead of meters, and uses meter resolution (and percentages) in table 2. We introduced different height abbreviations:
APL (above plain height), MSL (above mean sea level), AGL (above ground level, i.e., "above terrain height"). The mentioned percentages refer to ABL height APL. - appendix 1: we rewrote section 2.3.2 to mention the parcel method and refer to the appendix for a detailed/quantitative comparison between the methods. We chose to focus on the tracer method and keep the parcel method in the appendix.
Citation: https://doi.org/10.5194/egusphere-2025-4302-AC1 - The imposed wind profile is used only in the sensitivity experiment REALu10 (and now in FLATu10 as well) and is not intended to represent a realistic valley–ridge wind structure. Our main motivation was to introduce a non-zero, large-scale background wind within the boundary layer, with a magnitude comparable to the conditions documented by Chen et al. (2016, their Fig. 5). We set the initial wind to zero near the surface to avoid exciting strong orographic waves during model initialization. As the simulation evolves, there is downward momentum transport resulting in a non-zero mean wind within the BL and a shear layer across the entrainment zone.
-
RC2: 'Comment on egusphere-2025-4302', Anonymous Referee #2, 07 Oct 2025
The authors conduct semi-idealized LES experiments under dry conditions to compare three scenarios: a flat plateau (FLAT), realistic terrain (REAL), and realistic terrain with added upper-level wind (REALu10). They analyze boundary layer depth, turbulence structure, tracer dispersion, and vertical mixing. Key findings include:
Small-scale orography accelerates early CBL growth and leads to localized deepening (up to ~500 m over mountains). Wind shear organizes convection into roll vortices, enhancing mixing and further increasing CBL depth (up to 9.4 km ASL). Terrain and shear effects are roughly additive, with the deepest and most uniform CBL occurring in REALu10. A tracer-based method for diagnosing CBL height is introduced and validated against a parcel method.
The manuscript in its current form has minor weaknesses that must be addressed before it can be considered for publication. These primarily concern a lack of methodological clarity, and occasional overstatement of conclusions.
This study makes an original contribution by quantifying the role of small-scale orography and shear in driving extreme CBL depths over the Tibetan Plateau. I am curious, compared with the stability aloft, the fine-scale terrain and shear effects are still the firs-order factor for the extreme high CBL? Please add more sensitivity simulation to compare their respective importance.
The description of the Terrain Processing methodology is vague and omits critical details necessary for reproducibility and full assessment. The statement that the terrain was "tapered" and then "smoothed with a Gaussian filter (200 m)" is insufficient. What was the standard deviation of the Gaussian kernel? Was the filter applied once or iteratively? A figure showing the original SRTM data, the tapered terrain, and the final smoothed terrain used in the REAL case would be immensely helpful to understand the degree of modification and its potential impact.
Flow Decomposition: The description of the Reynolds averaging procedure is confusing. The authors use a 2D Gaussian filter on time-averaged fields. This is a spatial filter, not a classic ensemble average. The justification for this hybrid approach over a pure temporal or spatial average needs to be strengthened. Furthermore, the choice of a 500 m standard deviation and a 4 km support width seems arbitrary; Please give an explanation why these values were used. A sensitivity analysis or a stronger justification based on the dominant turbulent length scales would be beneficial.
Initial Conditions and Forcing: The use of an idealized sounding with a constant weak lapse rate above a 300 m mixed layer is a major simplification. The authors must more explicitly discuss the limitations this imposes, as the real atmosphere over the TP likely has multiple layers of varying stability that could significantly modulate entrainment and CBL growth.
Local vs. Domain-Averaged Impacts: While the local enhancement over the ridge (R-TOPO) is well-quantified (~500 m), the domain-averaged impact of orography (REAL vs. FLAT) is surprisingly small (~80 m). The authors should discuss this dichotomy more thoroughly. Does this imply that the overall impact of small-scale orography on the plateau-scale CBL is minor, but it creates strong local heterogeneity? The current narrative emphasizes the importance of orography, but the domain-mean results tell a more nuanced story.
The REAL terrain is described as featuring slopes up to 30° and ridges up to 1400 m. It would be helpful to contextualize this within the broader Tibetan Plateau to justify its representativeness.
While the idealized setup is appropriate for isolating mechanisms, the authors should more explicitly discuss how the exclusion of moisture, clouds, and synoptic variability might affect the generalizability of the results.
The authors mention that unresolved orography may bias CBL representation in models. A more quantitative discussion—e.g., how much CBL growth is underestimated in models that smooth terrain—would enhance the impact of the study.
The results section frequently references figures (e.g., Fig. 3, 4, 7) before the reader has any context for what these figures show. The narrative would be much clearer if the description of each figure was immediately followed by the interpretation of its results.
The abstract and conclusions are somewhat repetitive. The conclusions could be strengthened by synthesizing the findings into a broader conceptual model of how terrain and shear interact, rather than just restating the results.
Line 6, Small-scale orography substantially accelerates early growth: by midday the CBL in REAL is ∼80 m higher than in FLAT, 80m higher than FLAT is substantially? Compare surface flux and stability in the free air, is it still substantially?
Line 7, locally above the mountain it is ∼500 m deeper. This sentence is confusing, please rephrase it.
Line 7, This terrain-induced advantage narrows later in the day, what is narrows? Please rephrase this sentence.
Line 12, Under clear-sky conditions, the plateau’s CBL can exceed 9 km within a single day given strong surface heating and weak stability aloft. This conclusion has been approved by Chen` study. It may not be proper to be included in the abstract and taken as a conclusion of this study.
Line 18, One distinguishing feature of the TiP’s ABL is its extraordinary
depth. It`s better to cite the reference here: X., C. and Y., M., 2025. The unique atmospheric boundary layer over the Tibetan Plateau (in Chinese). Chin Sci Bull, 70: 4180–4187.
Line 37, it is better to further point out the weak stability in the lower troposphere is associated to the tropopause folds activity (Chen et al. 2011 ACP).
Chen, X., Ma, Y., Kelder, H., Su, Z. and Yang, K., 2011. On the behaviour of the tropopause folding events over the Tibetan Plateau. Atmos. Chem. Phys, 11: 5113-5122.
P3, L75: Specify what "REAL terrain" represents in terms of maximum height AGL and horizontal scales.
P4, L105: The vertical grid spacing is described twice with slightly different wording. Consolidate this into a single, clear description.
P9, eq. 4-6, The equations for TKE components are incorrectly formatted. w'^2 is repeated for all three components. This must be corrected to u'^2, v'^2, and w'^2.
Citation: https://doi.org/10.5194/egusphere-2025-4302-RC2 -
AC3: 'Reply on RC2', Ivan Bašić, 10 Dec 2025
We thank Reviewer 2 for their constructive and thoughtful comments. We appreciate the reviewer’s recognition of the manuscript’s strengths and their helpful suggestions for improving clarity and moderation of our conclusions. We have carefully addressed all points and revised the text accordingly.
==========
Responses
==========- The weak stability aloft is indeed the primary control on extreme CBL growth, as it sets the background conditions for deep dry convection (see energy balance model in Chen et al.). The present study focuses on how small-scale orography and wind shear further modulate this growth under otherwise identical, weak-stability conditions. In other words: small-scale orography and shear are the sensitivity test to the already favorable conditions for deep CBL growth.
In the manuscript, we mention this in section 2.1 and make it clearer overall. - We revised the terrain processing methodology (section 2.2), added Appendix B discussing super-gaussian filter, and added additional panel in Fig. 1 for more clarity.
- The 2D Gaussian filter.
It was applied to isolate sub-mesoscale variability in the time-averaged fields, serving as a spatial decomposition rather than a strict Reynolds average. The so-called "ensemble average" is then used. The filter parameters (σ = 500 m and support width = 4 km) were chosen deliberately to match the dominant turbulent length scales observed in the simulations, corresponding to the typical horizontal size of large thermals and roll vortices in the CBL.
Additional tests with smaller (σ = 250 m) and larger (σ = 1 km) smoothing lengths yielded consistent qualitative patterns, confirming that the decomposition is not sensitive to the exact parameter choice. We have clarified this rationale in the revised manuscript.
Why did we chose to include spatial-average as well as the time-average? With only the time averaging, the turbulent statistics are poor - but domain-means are not applicable due to orography. So we are choosing time averaging AND 2D Gaussian, to improve turbulent statistics. This approach is a "poor-man's" approximation to ensemble-average. - Initial conditions and forcing.
The weak lapse rate above the 300 m mixed layer was indeed chosen intentionally to reproduce the favorable thermodynamic conditions observed during deep boundary-layer development over the Tibetan Plateau. In particular, Chen et al. (2016) report weak free-tropospheric stability (near-dry-adiabatic lapse rates) during strong surface heating episodes, which we use as a representative reference.
We acknowledge that real soundings often exhibit multiple stable layers and inversions; however, simplified profile was chosen to isolate the effects of local orography and shear without introducing additional complexity from vertical layering. The goal was to conduct a sensitivity study, not reproduce observations accurately. We have clarified these limitations in the revised manuscript. - local vs. domain-averaged impacts.
You are exactly right and we tried to point it out in the manuscript, but it was not emphasized enough. What you observed reflects the fact that the orographic influence is spatially heterogeneous — it strongly amplifies local entrainment and turbulence over elevated slopes but averages out when integrated over regions that also include valleys and flat areas.
We have clarified in the revised results section and conclusions that subgrid orography mainly enhances spatial variability rather than substantially altering the domain-mean CBL height.
Also, we added a new Fig. 9 where we show how the boundary layer is actually less deep over orography - although may reach higher heights. This distinction is made clear (height above ground level vs. heigh above mean sea level). - terrain representativeness.
The chosen terrain represents a typical ridge–valley structure within the central Tibetan Plateau, with elevation differences of 1000–1500 m and slopes up to ~30°, consistent with regional SRTM data. The periodic boundary conditions ensure that the domain effectively repeats this representative mountain structure, rather than emphasizing any single extreme feature. The local context of the selected region is already illustrated in Fig. 1a, where the surrounding orography can be seen. We have added this clarification to the Methods section.
In the revised manuscript, the slopes go up to ~36°. - idealized setup.
We have expanded the discussion of limitations in the Conclusions to more explicitly address how the exclusion of moisture, clouds, and synoptic variability affects the generalizability of our results. As suggested, we now clarify that the objective of our idealized setup was to isolate the mechanisms by which orography and shear influence CBL development, and that these dynamical mechanisms can reasonably be generalized to other high-elevation or weakly stratified environments. We also note that moist convection and cloud processes would likely enhance vertical exchange and turbulence compared to our dry configuration, potentially amplifying some of the effects identified here. Finally, we specify that the synoptic background in our case corresponds to unusually weak stability, which is a key condition favoring deep CBL growth, and we point to future work needed to explore a wider range of synoptic regimes. - unresolved orography.
We have added a short quantitative statement in the discussion to emphasize this implication. Also, in section 3.4. we elaborate on this and clarify that models underestimate extreme BLs, and that relative ABL depth is smaller over the mountains (new Fig. 9). - figure referencing.
Thank you for pointing this out. We moved the figures so they appear earlier in the document, i.e., before they are referenced. - abstract and conclusions repetitive
Abstract is revised.
In conclusions: added future guidelines and also how can it be expanded. - lines 6 and 7.
We rephrased this part in the abstract to (also) include some ideas from the previous comment about local vs. domain-averaged impacts. - line 12.
We revised the abstract. - line 18.
Reference added. Thank you for the suggestion. - line 37.
The reference has been substituted. - line 75, 105, and equations 4-6.
All revised as suggested.
Citation: https://doi.org/10.5194/egusphere-2025-4302-AC3 - The weak stability aloft is indeed the primary control on extreme CBL growth, as it sets the background conditions for deep dry convection (see energy balance model in Chen et al.). The present study focuses on how small-scale orography and wind shear further modulate this growth under otherwise identical, weak-stability conditions. In other words: small-scale orography and shear are the sensitivity test to the already favorable conditions for deep CBL growth.
-
AC3: 'Reply on RC2', Ivan Bašić, 10 Dec 2025
-
RC3: 'Comment on egusphere-2025-4302', Anonymous Referee #3, 07 Oct 2025
This manuscript investigates the impact that small-scale orography features have on the evolution and structure of the convective boundary layer over the Tibetan Plateau. The author used an LES model (CM1) and ran three idealized simulations for a special case using three different configurations that differ in the small-scale orography features included and in the presence/absence of upper-level wind. The results show these small-scale features and shear effects enhance the CBL growth by up to 15%. The paper is well designed, clearly states the objectives and methodology, and the results are properly presented and discussed. I recommend the paper for publication, although some minor comments should be addressed before that.
Minor corrections
- Line 4: replace “a upper-level” by “an upper-level”.
- Line 8: please explain what LT means (first time that appears in the text).
- Line 16: please explain what the “TiB” abbreviation means.
- Line 19: same for the abbreviation “CBL”.
- Line 94 to 95: sentence repeated (already explained in line 91).
- Line 96: please include a brief explanation of what semi-slip condition means.
- Line 116: I think “FLAT” is a bit confusing cause this simulation is not flat itself, but does not include small-scale features. Please consider changing the name, like NOSSO or something that states more clearly how the orography is represented in that simulation.
- Section “Experiments and Data”: could the authors include whether the simulations were run with a spinup period?
- Lines 145-147: these two sentences say the same in a slightly different way. Please consider keeping just one to avoid repetition.
- Line 153: include the symbol q after potential temperature, as this symbol will be used in the caption of several figures throughout the manuscript.
- Figure 2: this figure is just briefly discussed in lines 154 and 155. I would suggest moving it to the Appendix.
- Line 157: add “w” after vertical velocity as this will be used in the caption of the following figures.
- Figure 3: in the caption, the authors mentioned that the vertical cross-sections are calculated for y=0. To improve the clarity of the manuscript, I suggest including a line in figure 1 showing where this cross-section is located.
In the caption of this figure, there is a reference to 2.32.3.2. Please correct. - Line 182: the authors say they also analysed the instantaneous vertical velocity at lower levels and that those results are not shown. Perhaps the one at 500 m AGL could be included in the Appendix.
- Line 189 and 190: w’2 should be replaced by u’2 and v’2, respectively.
- Figure 7: please indicate the units used and explain the meaning of the lines shown.
- Line 202: FLAT should be replaced by R-FLAT if I am not mistaken.
- Figure 8: if I am not mistaken, y-axis is showing the altitude of the CBL height and not the surface-based tracer concentration. Please clarify this.
- Line 281: the Appendix is never mentioned throughout the manuscript. Perhaps, you could refer to it in section 2.3.2 when talking about the method used to diagnose the boundary layer height.
Citation: https://doi.org/10.5194/egusphere-2025-4302-RC3 -
AC4: 'Reply on RC3', Ivan Bašić, 10 Dec 2025
We thank Reviewer 3 for the comments and corrections. The attention to detail is really helpful and it improved the manuscript.
==========
Responses
==========- line 4,... line 96: revised as suggested.
- line 116:
When introducing FLAT, we write: FLAT (no local orography). - spinup period.
The first 2 hours are considered spinup. They are discarded as the model spinup to achieve a reasonable physical atmospheric state. We added a sentence to section 2.1 regarding this. - lines 145-147: revised.
- line 153: revised
- Figure 2.
The authors consider Figure 2 to be an important part of the paper clarity and reading flow and we decide to keep it as-is - but we expanded the discussion around it in Section 3.1. - line 157: revised
- Figure 3.
The authors think adding the horizontal line is unnecessary, and that the one added in the top row of Fig. 3 serves this purpose. However, if the reviewer insists, we would add the horizontal line.
Caption is revised. - line 182.
We added Fig. C1 (in the appendix) with four panels showing two heights AGL: 250 m & 500 m. - line 189-190: revised
- Figure 7.
We added the units and we revised the figure so that the upper-level tracer is dimensionless (percentages) rather than actual values, since it is more helpful for a qualitative read. - line 202: the authors carefully checked the line, and "FLAT" is correctly placed.
- figure 8: revised.
- line 281: revised as suggested.
Citation: https://doi.org/10.5194/egusphere-2025-4302-AC4
-
RC4: 'well-written manuscript, but more analysis on TKE budget neeed', Anonymous Referee #4, 08 Oct 2025
Dear authors
As a first note, I want to inform you that I already read through the three other review statements. In general, I agree with all of them, and henceforth, my comments might be less than a typical review, but please consider them as equally relevant.
General comments
This is a clear, well-written manuscript exploring the reasons for a very high-ranging boundary layer over the Tibetan Plateau. Given the Tibetan Plateau's role in the global climate system, this research is relevant for the community. The authors employ the CM1 model for idealized simulations and analyze the boundary-layer structure in simulations with/without orography, and imposing different background wind conditions. The manuscript is well-written, but additional analysis (TKE budget) and possibly one more simulation (FLAT/u10m) are necessary to strengthen the conclusions of the authors and to make the MS suitable for publication in ACP.
Major comments
You identified shear-induced turbulence as one of the major reasons for the high-ranging ABL over your plateau. However, although you mention shear throughout the manuscript, you never show its actual quantity, e.g., by calculating the shear production term in the TKE budget equation compared to buoyancy forcing. Creating a figure and the accompanying text of the TKE budget will strengthen your manuscript massively.As a second point, you talk about 'terrain forcing'. I understand that this conclusion is drawn from a comparison of flat vs. complex terrain. However, if I understood it right, the FLAT simulation is a sub-domain of the domain with real toporgaphy.
1) How can you make sure that the FLAT subdomain is not affected by a plain to mountain flow towards your plateau?
2) As a second point, it would be also advisable to add a simulation/analysis of a flat-terrain, but shear-induced simulation, i.e., FLATu10, so that you can make a direct comparison with the plateau simulation - and - especially relevant - to isolate the 'terrain' and 'shear' effects accodingly.Minor comments
As other referees already pointed out minor formulation or speeling errors, I would only want to encourage the authors to avoid statements like 'typical development' or 'expected behaviour' without citing the relevant literature. Not all readers are aware of how boundary-layer development over orography is 'supposed to look like'. Furthermore, the last paragraph of the introduction might better fit in the methods Section.Therefore, I recommend major revisions for the manuscript and look forward to the revised version.
Best regards
Citation: https://doi.org/10.5194/egusphere-2025-4302-RC4 -
AC5: 'Reply on RC4', Ivan Bašić, 10 Dec 2025
Thank you for your valuable comments and insights, and very helpful suggestions. We address each of them carefully.
===============
Major comments
===============- Shear production.
We add additional panel to Fig. 5, which was previously just a resolved TKE profile at 15 LT. Now it includes the shear and buoyancy production term at 15 LT as well. We expand the accompanying text in section 3.3. - Terrain forcing.
The FLAT simulation is a stand-alone simulation with no real topography. There is a subdomain called R-FLAT and it is only analysed in Fig. 8.
Answering "1)": there is no plain-to-mountain flow (see Fig. 1d). However, there is plain-to-mountain flow in R-FLAT, but it is intentional since it is a subdomain and analyzed as such.
Answering "2)": we added an additional simulation to the study: FLATu10 to do exactly what the reviewer suggested, and it helped us show our concluding points more clearly.
===============
Minor comments
===============- The authors added the relevant literature (Stull, 1988) and rephrased the sentence that mentions the typical TKE profile in the late afternoon to make the statement more clear.
- Last paragraph of the introduction was rewritten to only contain an overview of what was done with no details. The details were moved to the Methods section, like the reviewer suggested.
Citation: https://doi.org/10.5194/egusphere-2025-4302-AC5 - Shear production.
-
AC5: 'Reply on RC4', Ivan Bašić, 10 Dec 2025
-
AC2: 'General note to the referees', Ivan Bašić, 10 Dec 2025
Dear referees,The manuscript has undergone major revisions, particularly in the Methods and Results sections. The main changes are:
- Addition of FLATu10: A new experiment (FLATu10) was added to better isolate the impacts of orography and wind shear.
- Background wind in FLAT and REAL: A weak background wind was introduced in FLAT and REAL (constant profile of 1 m/s) to break the W–E symmetry. This modification was successful (see Figs. 3, 6, and 7).
- Terrain smoothing settings: The standard deviation σ of the 2D Gaussian filter applied to the terrain was reduced from 200 m to 75 m (a factor of 4 to 1.5 grid spacings) to retain more small-scale terrain features. While the overall results do not change, the differences are noticeable in 99th percentiles, which reinforces our main conclusions.
- New Appendix B and additional figure: Appendix B was added to describe the super-Gaussian tapering of the lateral boundaries, and Fig. C1 now shows w' perturbations at lower levels, which were previously only mentioned in the text.
- New ABL–terrain diagnostic (Fig. 9): Figure 9 was added to show ABL depth (km AGL) and ABL height (km MSL) as a function of underlying terrain elevation, providing additional insight into the influence of small-scale orography.
- Greater emphasis on the upper tail of the ABL height distribution: we highlighted the behaviour of the upper tail (99th percentile) of the ABL height distribution (Fig. 8; Table 2; Section 3.4; Conclusions). This clarifies how small-scale orography disproportionately affects the deepest convective columns.
Kind regards,
The authorsCitation: https://doi.org/10.5194/egusphere-2025-4302-AC2
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 2,134 | 156 | 52 | 2,342 | 97 | 92 |
- HTML: 2,134
- PDF: 156
- XML: 52
- Total: 2,342
- BibTeX: 97
- EndNote: 92
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Harshwardhan Jadhav
Jaydeep Singh
Juerg Schmidli
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(17854 KB) - Metadata XML
Review in connection with the manuscript entitled
Impact of small-scale orography on deep boundary layer evolution and structure over the Tibetan Plateau
by Basic et al.
Manuscript: https://doi.org/10.5194/egusphere-2025-4302
General
In this paper the authors use high-resolution LES (CM1) in idealized experiments to study the impact of ‘subgrid-scale orography’ (subgrid to typical global NWP or regional climate models) on the development of extremely deep convective boundary layers on the Tibetan Plateau. It is found that beyond the earlier identified conditions of weak stratification aloft and not too weak surface heating, ‘local orography’ indeed has an enhancing effect on CBL growth – thus leading to order 10% higher CBL. Overall, the paper is well designed (numerical experiment) and the results are presented in a consistent manner. I have, however, two ‘major comments’ addressing some details of the experiment. They are called ‘major’ because they cannot be attributed to a line (as the minor comments) – but will easily be addressed. In this sense, I think , the paper can be published after having addressed the minor comments.
Major comment
Minor comments
l.16 ‘TiP’: the abbreviation is introduced in the abstract – but this usually not counted. So, please introduce the abbreviation here as well.
l.70 ‘FLAT (no orography)’: this may be a matter of definition, but I consider the TiP to be part of ‘orography’, too. So, maybe the authors want to consider to re-label this to locally flat (‘no local orography’, no ‘subgrid-scale orography ‘) or alike.
l.91 this ‘delta20 m’ notation seems to be odd. Why not ‘the vertical grid spacing is set to 20 m below….’? Several occurrences on this and the next line.
l.94 ‘the vertical grid spacing…’: this is repeated (and this time the ‘delta’ is already removed….)
l.96 what is a semi-slip condition? Please explain.
l.98 ‘idealized sounding’: please specify whether this only refers to temperature, or also includes humidity and wind speed (the u10 simulation suggests that the reference has no wind?)
l.103 ‘…starting a 09LT….’: is this the start of the simulation? Or is there a spin-up period considered? Please specify.
Tab 1 ‘Flat orography’ seems to be a contradiction. In the text it is referred to as ‘no orography’, which also seems to be counter intuitive...(see comment to l.70). Again, I suggest to consider ‘no subgrid scale orography’, or ‘elevated plain’ or similar).
l.139 ‘to quantify’ is probably not appropriate here – consider ‘to diagnose’ or ‘to determine’ instead.
Fig.3, caption: what does ‘see 2.32.3.2‘ refer to? Also, the black contour line should be explained, as 800 m above plateau level (i.e., 4200 m ALS)
Fig.3 In the caption it says ‚at 1800 m AGL (6 km ASL)‘: for the ‚REAL‘ simulations the equivalence of these two statements is not given. Please specify, which applies.
Fig. 4, caption: must specify the time, when the instantaneous fields are obtained.
Eqs (4) and (5): the lhs must be u’^2, v’^2 respectively. Furthermore, if using the according to (1) and using (2) the ‘difference between the time-averaged second and first moments’ is not obvious. I suggest to explicitly derive this (maybe in an appendix). However, in the nomenclature of the present work, it would, correspond to the resolved variance. This should be added.
Fig.7 you have introduced the tracers as ‘1’ and ‘2’ – and here, they are referred to as ‘surface-based’ and ‘upper-level, respectively. I suggest to introduce this convention where the tracers are introduced. Also, in the caption of Fig. 7, the units of the isolines must be specified.
Fig.8, caption: ‘horizontal distribution...’: if I understand the accompanying text correctly, these are the horizontal statistics (median, interquartile range) of the diagnosed mixing height (or CBL height) based on the surface emitted tracer and the elevated tracer (and not: tracer concentrations). I furthermore cannot understand how ‘the mean concentration’ can be marked (with an x) in a height domain (time) diagram. Please explain.
l.249 ‘…by 383 m…’: I don’t think that ‘meter resolution’ here is appropriate (I much more like the ‘about 10%’, even if in the summary then it should be indicated percent of what). Putting into context could also be done by expressing the change in percent of the terrain height.
Appendix 1: this appendix is never mentioned in the manuscript, nor is the parcel method. So, either the appendix can be removed from the ms, or the parcel method is included into the analysis.