The introduction makes a detailed analysis of the state of the art in the study area. The supercell database is reliable and accurately described. The methodology is clear and robust. The results are coherent with previous studies, and they are properly accompanied with tables and figures. The conclusions are consistent with the results and discussion previously carried out. As a result, I recommend the publication of the manuscript after minor corrections that I propose in the following lines.

Major comments

Section 3.2. I would not refer to mesoscale settings by using the ERA5 database. I encourage the authors to change the title of this subsection, as well as all mention of mesoscale throughout the paper.

Minor comments

Line 32: “a convectives storm”. Write this sentence in a proper English.

Lines 51 to 61: This paragraph is an enumeration of what other authors did in the past. Try to improve the writing, linking some sentences with others, to engage the reader.

Line 97: Missing dot between “(ECMWF) It”.

Line 306: “907 tornadoes” instead of “907 tornado”.

The introduction and discussion of the results is quite detailed, but the authors can consider to complete them with the following research:

E. Gascón, A. Merino, J.L. Sánchez, S. Fernández-González, E. García-Ortega, L. López, L. Hermida. 2015. Spatial distribution of thermodynamic conditions of severe storms in southwestern Europe, Atmospheric Research, Volumes 164–165, 194-209, https://doi.org/10.1016/j.atmosres.2015.05.012.

Title: Supercell Convective Environments in Spain based on ERA5: Hail and Non-Hail Differences

Authors: Calvo-Sancho et al.

RECOMMENDATION: Rejection

The purpose of the paper is to compare the environments conducive to supercells with and without hail to identify differences. The analysis is performed over Spain using ERA5 reanalyses over 10-years. The study is potentially interesting, but I do not think the quality of the paper in the present version is sufficient to justify the publication in a high-quality journal. I have serious concerns about the methodology, the analysis of the results, the quality of the figures. The paper is difficult to read and appears mainly as a very preliminary, disorganized draft that would have still needed a lot of revision before submission. In the following, I would not focus on the style since there is a lot to be done on the content. So, I recommend a complete rearrangement of the paper and improved analysis, although I encourage resubmission.

Introduction:

• the focus of the paper is on the comparison between hail and non-hail supercells, so you should remain focused on that also in the introduction. Also, the studies you mention in the introduction are almost exclusively focused on Spain, while you should extend your comparison of hail vs non-hail supercells to the whole Mediterranean and possibly other regions;
• Line 45: “supercells in Europe tend to be smaller, both horizontally and vertically, than those formed in the US”: do the papers you refer to include quantitative estimation of the reduction in horizontal and vertical extent? I think this is a difficult task to assess, so I would be curious if there are some statistics supporting this statement. Similar considerations apply to the "reduced rotation and shorter life spans": is there any statistics to support your sentence?

Section 2:

• Has the supercell dataset been validated somehow? For example, did you make a comparison with the hail occurrences as reported in ESWD or in local datasets?
• Line 83: you mentioned earlier that the medium-high confidence events are detected in radar images but without direct observation; here, you mention that thanks to volunteers, 20.5% of the medium-high confidence supercells were confirmed by two-dimensional radar images. Sorry, but I am confused.
• Line 143: I think it is interesting that the only parameter changing with time is CIN, possibly as a consequence of the change in the environmental conditions after convection is triggered.
• Line 146-147: I disagree with this point. Once convection is triggered, the environment should be "contaminated" by the vertical redistribution of temperature consequent to the vertical motion, thus the profile at tc would be less representative of the environment conducive to supercell development than that at the earlier stage.
• Line 148-149: I do not agree that the information related to WS06 is more important than that on MUCIN. The fact that MUCIN is different reveals that the environment has substantial differences between tc and t0, i.e. before and after convection is triggered (you wrote that other buoyancy terms can be evaluated, but I do not see which ones you consider here).

Section 3:

• Line 164: I do not understand why the eastern half of Spain should be special from the point of view of upper-level forcing for ascent.
• Line 170: in other Mediterranean areas the peak of hailstorms occurs in June (e.g., Manzato, 2012), due, as expected, to a combination of strong diabatic heating and cold air intrusions, more frequent in late spring-early summer. Why does Spain behave differently?
• Figures 3, 4, 5: To highlight the differences, I suggest showing the SP-HAIL fields and the differences compared to the SP-NONHAIL fields. In the present version, it is difficult to detect the rather small differences. In addition, in Figure 5: contour lines are very difficult to identify, differences are not clear, coastlines can be hardly identified.
• Line 197: do you mean short or small in amplitude?
• Line 205 and elsewhere: moisture, not moist;
• Line 208 and elsewhere: easterly winds, not eastern;
• Line 209-210: I would rather say that the difference in DWPT is mainly a consequence of the different dominant seasons in the two supercell datasets.
• Lines 211-213: “The high elevations reduce the role of convective inhibition, which is also met by the convergence of southwestern and eastern surface winds”: what do you mean??? do you mean that the orography forces the air parcels to be lifted above the LFC?
• Line 223: “This maxima omega area matches with positive Q-vector divergence values … and convergence of Q-vectors”: I do not understand: do you mean that maxima omega values are superimposed with both divergence and convergence areas???
• Line 225: “higher values of maxima omega in SP-NONHAIL at 850-500 hPa”: why do you consider in the following analysis only the maxima omega vertical velocity at 700-400 hPa thickness in SP-HAIL and not the maxima omega in SP-NONHAIL at 850-500 hPa?
• Line 227: wind convergence does not enhance and reinforce convection, rather favors triggering.
• Line 231-232: I am very confused, I see values of order 30 m/s, never below 20 m/s, in Fig. 5.
• Line 245: 90-th percentile with respect to what?
• Line 247: what do you mean with “a better buoyancy distribution”?
• Line 248-249: I do not see where the value of CAPE is reported, and where you show that the CAPE values are larger at t0 than at tc;
• Line 250: “The CIN in SP-HAIL increases from t0 to tc”: where do you show this increase? also, it is very hard to physically understand why CIN increases: should not the convection remove progressively the inhibition?
• Line 252: -137 J/kg is a rather extreme value for CIN, I do not believe that convection can develop even in the presence of mountains with such a value; conversely, a value for about -50 J/Kg as reported by Taszarek et al. (2020b) (Line 317) appears more reasonable.
• Line 254-255: “a higher LCL is related to the width of the deep convective updraft, resulting in a wider, deeper, and faster vertical velocity”: I do not understand how general this result is and the physical reasoning for that;
• Line 257: why is it relevant to have large WS values “above” the updraft height?
• Line 260: “The evolution from t0 to tc depicts a reduction in WS for SPHAIL”: where do you show this point?
• Line 261: what does “contrary to the SP-NONHAIL episodes” refer to?
• Line 266: “Different distributions can be seen in Figure 7”: I would say this is not relevant, it is rather a consequence of the different dominant seasons in the two categories.
• Line 266: what do you mean with “bimodal distribution”? I do not see it in Figure 7.
• Lines 272-276: I think the differences in humidity are mainly due to the different seasons prevailing in the two categories and not to the different wind features;
• Line 284: what do you mean “with a lower amplitude”?
• Lines 287-288: I would rather state that MLCAPE is very close to SBCAPE.
• Line 304: you cannot compare values in high-resolution models with those in reanalyses.
• Line 305-307: “the CAPE values found in our study would correspond with those for tornadic storms … finding SBCAPE values higher than 400 J kg-1 in tornadic storms”: here you find that 75% are below 400 J/kg, so they do not correspond.
• Figures 8, 9, 10: what time do the figures refer to?
• Lines 319-324: the causes you address for the high CIN would be relevant in case you consider soundings at times distant from the development of the cell, while it is very strange that you have such a high CIN during or in the proximity of convection.
• Line 336 “the median MLLCL in SP-HAIL events is greater”: this may be due to the presence of the mountains.
• Line 337: how do you interpret physically the higher LFC for SP-HAIL? I would rather expect that a lower LFC would favor deeper updrafts and then stronger hail formation!
• Line 339: values of MLLFC higher than 2000-3000 m appear extremely high (you even obtain values of 5000 J/kg!): how do you explain them?
• Lines 350-355: Figure 10 shows the opposite compared to what you wrote, i.e. wind shear is higher for SP-NONHAIL.
• Line 363: about SHR01 “the median value of SP-NONHAIL is higher than SP-HAIL”: this appears counterintuitive: any explanation for that?

REFERENCE: Manzato, A., 2012: Hail in NE Italy: Climatology and bivariate analysis with the sounding-dervied indices, J. Appl. Met. Clim.,  51, 449-467.

This work focuses on environmental evaluation of supercell thunderstorms across Spain. Authors combine supercells producing and non-producing hail with proximal ERA5 environments and discuss several convective parameters and accompanying synoptic-scale patterns. Authors divide results into the early and mature stage of the supercell thunderstorm. While this is an interesting work addressing a niche of very much needed studies focusing on supercell thunderstorms across Europe, it requires major revisions to meet certain publishing quality. My most important concern is that authors have a tendency to make statements that are not scientifically relevant while in other instances speculative. My feeling is that authors sometimes ‘overdo’ interpretation of their results and do not entirely understand the subtle difference between the mesoscale environment derived from a coarse-grid ERA5 and local storm-scale features such as convective updraft that can be only resolved with high-resolution convective-allowing models. A good example is L387 where authors write „Omega vertical velocity reveals that the SP-HAIL's updraft is higher” or „sounding composites show large wind values in upper-levels, which may favor wind divergence at the upper troposphere and deep-moist convection„. Manuscript contains several such statements that need to be revised before the article can be accepted. There are also awkward sentence constructions (e.g. „a mechanical trigger to force the mechanism that initiates convection„) that in certain places make it difficult to understand the meaning of the sentence. Thus, further language proof-reading of the manuscript is required.

I also have a feeling that analysis of differences between t0 and tc is a redundant part of this study as it doesn’t introduce important findings. This is especially strange given that authors try to find differences in small details over small distances between t0 and tc, but at the same time they average their profiles to 9x9 grids and do not benefit from 0.25 deg resolution of ERA5. Trying to evaluate subtle differences among closely located t0 and tc for large synoptic-scale features at figures 3, 4 and 5 is even less scientifically relevant. While I like the concept of dividing supercells into hail producing and non-producing events, I am just skeptical whether division into t0 and tc is worth all the attention authors devote in this study. This is not a major issue and I leave the decision regarding incusion/exclusion of this part to authors. At the end of the day it is their decision what and how they want to present in their work. However, for future studies with this dataset, instead of ERA5 with 9x9 grid averaging, a convective-allowing high-resolution simulation would be likely more appropriate to evaluate different stages of the supercell lifecycle at t0 and tc and investigate the influence of ambient environment and local orographical features.

In this work, I think that division of results into high-CAPE and high-shear events would be probably more interesting and scientifically important in the context of other similiar work that has been done for Europe (compared to t0 and tc approach). It is well known that European severe storms are mostly driven by strong kinematics and in lower degree by high instability, which is also a case for supercells.   

Minor comments:

L15: Suggest changing to „the synoptic configurations and proximity atmospheric profiles related to the supercell events”.

L19-L21: Awkward sentence construction, please rewrite for clarity. Perhaps splitting this sentence into two can help.

L27: Suggest changing „life” to „lifecycle”

L35: I am not entirely sure I can agree with this sentence and the phrase „easily detected”. Supercell detection in Europe is generally not easy if high-quality Doppler radar data is not available (like in the U.S.). I am also not sure how a mesocyclone (which is a core definition of the supercell thunderstorm) can be detected by lightning data. In the majority of instances we can only suspect that supercell thunderstorms developed based on its morphological features, but only a small fraction of these events can be captured by nearby Doppler radar velocity products that provide ultimate confirmation of the mesocyclone. I suggest authors reword and soften this sentence or remove it.

L43: No need to use „observational” ahead of „reports”. Authors may consider using „severe weather reports” instead.

L45: Is there any scientific proof that they are indeed smaller? Authors speculate that it is due to oroghraphy and land-sea interactions, but is it really the case? What about big supercells in Nebraska or Southeastern U.S. along the coast of GOM? Is there any scientific proof showing that orography acts to reduce size of the supercells? Perhaps weaker supercells in Spain are rather due to smaller CAPE and WS / less favorable wind profile compared to their U.S. equivalents. I suggest rewording.

L59: Which „other regions of the world”? Please be more specific.

L95-101: Did authors also use surface data in addition to pressure levels, and eliminate all pressure levels falling below orography for the purposes of parameter calculations? This information should be included in this paragraph. Also, which software was used to calculate convective parameters. SHARpy, MetPy, other, or your own scripts? Did you also consider that some of the proximity profiles may be contaminated by the convection ongoing in ERA5? Did you use convective precipitation threshold equalling 0mm to eliminate such profiles? This might be an approach worth considering in potential future studies to make sure evaluated profiles are pre-convective.

108-111: I am not sure if that was a good idea. In this way authors do not benefit from the superior (compared to other reanalyses) resolution of ERA5. This averaging can have an impact on areas with complex orography and result in the loss of important details. Did authors try to reproduce their results without a 9x9 grid averaging approach? Were these results much different?

L124: What authors mean by „The 2-meter temperature (T2M) and dew-point (DWPT) are computed”. In which aspect T2M and DWPT required computations? To avoid using a word „computed” authors can reword into something like „We selected the 2-meter temperature (…)”.

L126: What depth was used for calculating mixed-layer?

L149: I am not sure if I understand what authors mean by „This variable is much more interesting than MUCIN, as there are other buoyancy terms which can be evaluated”.

L205: Change „moist” to „moisture”.

L244: Why do authors think that 90th percentile of MU_CAPE would indicate „largest and severe supercells”? Suggest rewording to „of the supercells developing in highly unstable environments”. Instead of providing mean skew-t profiles divided into t0 and tc it could be potentially interesting to provide also mean skew-t profiles for 90th percentile of WS events as high WS is a major contributor to severe storms in Europe compared to instability that is often limited.

L256: What exactly „helps to organize convection”? Please rewrite for clarity.

L259: „Also, the sounding composites show large wind values in upper-levels (< 400 hPa), which may favor wind divergence at the upper troposphere and deep-moist convection” – wind values from single profile cannot be used to determine upper-tropospheric divergence and deep-moist convection. It is a spatial pattern of the pressure field that allows to determine divergence and potential areas for the large-scale lift that may trigger deep moist convection. Please rewrite.

L260: „The evolution from t0 to tc depicts a reduction in WS for SP- HAIL, which is mainly denoted in the wind speed and not in the rotation” – I do not understand what authors mean by „and not in the rotation”. The degree of veering in the vertical wind profile?

L272: „These differences are mainly originated in the low-level wind flows.” - awkward sentence construction, please rewrite for clarity.

L280: CAPE can be a useful predictor but only with the combination of vertical wind shear. Over the tropics there is plenty of CAPE but rarely any supercell or large hail due to weak WS.

L286: How CAPE can be dependent on the orography? Please be more specific. Over northern Great Plains CAPE can reach as high as 9000 J/kg over higher elevation in Nebraska.

L288: Larger compared to what?

L311-L312: I believe this sentence is inaccurate. It is not only an ERA5-related issue but nearly every reanalysis (or NWP dataset) and is related to limited vertical resolution of available levels. applied convective parameterizations and convective contamination. Given that authors used less numerous pressure levels (instead of more frequent sigma levels), CIN values are expected to be less accurate as well. However, as shown in other studies, compared to other reanalyses ERA5 still performs better for CIN (e.g. table 2 in https://doi.org/10.1175/JCLI-D-20-0484.1). I suggest to soften this sentence and reword it to something like: „It is well known that due to limited vertical resolution reanalyses do not represent capping inversions very well”.

L321 Airmass advections from NW Africa and development of elevated mixed-layers can be also another reason for higher CIN across Spain and W part of Mediterranean compared to other parts of Europe.

L325: „a mechanical trigger to force the mechanism that initiates convection„ – awkward sentence construction, please rewrite for clarity.

L357: Helicity or rather storm-relative helicity?

L359: Period missing before „Environments”. Also, this sentence has an awkward construction, please rewrite for clarity.

L387: „Omega vertical velocity reveals that the SP-HAIL's updraft is higher” – ERA5 omega vertical velocity derived from 0.25 deg grid and averaged by authors to 9x9 matrix surely does not tell anything about local storm-scale convective updraft. 

Figure 3, 4 and 5: Text that is at the top of each figure and x and y axis is too small and impossible to read.

Caption to figure 6. 90th percentile of what? Please be more specific in the figure caption.