EGUsphere-2022-877 Review

Seasonal Controls on Isolated Convective Storm Drafts, Precipitation Intensity, and Life Cycle as Observed During GoAmazon2014/5

Giangrande, Biscaro, and Peters

Recommendation: Accept with minor revisions

General Remarks: This study examined the variability in cloud features that develop and propagate over DOE-ARM GoAmazon2014/5 campaign observation site (T3) and SIPAM radar system. The cell features are tracked with well cited track methodology with a few alterations. While numerous past studies have investigated the seasonal-composite cloud properties, this study analyzes the cell life cycle using radar cell tracking with the addition of complimenting observations datasets. The results of this study indicate that there was a difference in the mean lifetime of cells that form in the wet and dry seasons. The DCC updrafts were found to be more intense during the dry season and generally DCCs reached peak intensity earlier in their life cycle. The usage of the simplified updraft model was an appropriate addition to further investigate the physical reasons for differences in the draft characteristics. The physical explanations describing the results were sound.

This manuscript was well written and has significant findings regarding deep convective cell development during the wet and dry season in the Amazon. Overall, this manuscript is of scientific value because it characterizes the life cycle of convective cells in various environments, analyzes the evolution of draft properties, and uses profile-based vertical air velocity information to further the understanding of deep convective cell dynamics. It is recommended that this manuscript be accepted with minor revisions.

Major Comments:

1. Is it possible to use statistical resampling methods to increase the sample size of your cells in the wet and dry seasons?
2. With the criteria used to define a cell of interest in this study, how many cells were removed for failing one requirement?
3. How is echo top height defined? Is it 18 dBZ or another threshold?
4. Is the choice of 250 retrievals (line 115) for the CFADs based on previous testing or a common approach?
5. If you replicated this approach using satellite measurements, would the cell counts be similar?
6. Would sub-setting the DCC events during the wet season (dry season) further by additional environmental conditions lead to any new findings about draft characteristics?
7. The introduction describes why understanding DCCs is important for climate model improvement and potential changes in the cloud population with potential climate change. There was no connection from the findings of the study back to these broader impacts.

Minor Comments:

1. Line 149: 40 km2 -- km²
2. Line 190: Not sure if the text needs to include specifics on how to interpret the lines of the figure.
3. Line 200: Could you include the specific time that there is a more rapid transition to deeper convection or annotate it in the figure?
4. Lines 208-214: The arguments for 25 dBZ and 35 dBZ discussion were different but the lines on Figure 3 looked very similar. Perhaps a difference plot will help illustrate the argument.
5. Line 226: Perhaps reference the specific values being discussed to guide the reader.
6. Line 234: An explanation of the meaning of light rain / periphery may be needed. How is intensity inferred from Figure 4?
7. Line 295: To later ETH-- wording is unclear
8. Line 323: Is the choice of 100 parcels based on previous testing or statistics, or computational limitations?
9. Is there much variability in the fractional entrainment rate of tropical convection? If a range of entrainment rates are utilized, are the results similar?

Figures:

10. In the figures it might be useful to include the sample numbers of wet season and dry season cells in the titles or legend.
11. Figure 2 subpanel d: Planet--Planetary
12. Figure 8 (6, 7): The font sizes of the axis labels, ticks, and color bars are very small
13. Figure 8: There is a way in python to ignore that panel h, unless it is empty for a reason? Perhaps in the figure caption explain why it is empty.
14. Figure 9: Color bars for the plots or specify the values of contour values
15. A figure showing the methodology of the tracking method would be a good addition to visual the steps of the tracking method.

Using the observational data collected during the GOAmazon campaign, this work characterizes a certain type of deep convective events over the Amazonnear the city of Manaus. Compared with previous studies---many of which contributed by the lead author---this manuscript focuses on the isolated, diurnally forced local events peaking in the afternoon that are different from the nighttime storms originating from elsewhere. The statistics synthesized here are useful as a background/benchmark for model comparisons. I think this manuscript is suitable for the ACP; I only have a few comments which should be easy to address.

Major comments:

1) The expressions “melting level” and “aloft” are used multiple times in the manuscript. The actual height of the melting level is not explicitly called out; it can be identified through the abrupt transition in the draft velocity or frequency in some of the figure panels, but not so obvious in some of the other panels. Similarly, for aloft, it’s probably better to give a more precise height range (including in the abstract).

2) Figure 2b shows no/little clouds in the mid-and lower-troposphere around 1600 LT, and the text in L199-201 states that this is caused by a rapid transition to deep convection in the dry season. I am not sure why the mid-and low-level clouds would disappear when a rapid transition occurred. In the dry season, Fig.4f indicates that the number of events hitting the T3 site drop to zero at some point; Is the absence of mid-and low-level clouds related to the small sample size?

3) I’d like to suggest adding grid lines to Figs. 4-8, or at least adding the w=0 line, making the figures easier to read.

4) I’ve looked at somelong-time (1-year) high-resolution (dx~1km) CRM output and noted that there wasn’t any buoyant plume that can be distinguished from the environment. Although buoyant plumes (and bubbles) have been important conceptually and have been used in CRM studies to initiate convection, gravity waves seem to be very efficient to diffuse the horizontal buoyancy gradient in the free troposphere (at least when f is small). This observation made me reconsider the use of traditional plume models in general. I am not disputing the fact that plume computations can be informative as used here; I am glad to see that some of the new elements developed by Peters and colleagues have been incorporated into the model; I acknowledge that there isn’t a valid alternative of the plume approach (except for the computationally expansive CRM/LES options). Still, I am interested in what the authors think regarding this point. In simulations for shallow convection we do see plumes/bubbles. Is there any evidence suggesting that isolated storms driven by diurnal forcing behave like rising buoyant bubbles?

Minor comments:

L99: AGL?

L105: What does “toward relative updrafts” mean?

L126: Is the lowest 1 km chosen according to the PBL height shown in Fig. 2d?

L155-156: Are these numbers of events different for Z>25dBZ vs 35dBZ? It may be handy to have these numbers repeated in Tables 1, 2 as well as Fig. 4f (they are kind of buried in the text).

L157-162: It’s probably better to call out in the beginning of L157 that these numbers are estimated using the Z>25dBZ threshold. L161 seems to be saying that the numbers for Z>35dBZ are similarto those for 25dBZ, contradicting to the numbers 90/30 minutes being different. 

L193-195: The reduction of dry season cloud cover is also associated with a slightly greater increase in surface temperature as in Fig. 2c. (To double check: Is this surface temperature or surface air temperature?)

L269: Bardakov et al. (2022; JGR Atmosphere) recently noted in a set of CRM experiments that air parcels in downdrafts were mostly from mid-troposphere, which may be relevant. Another related point: There’s a paper by David Romps that studied the lifetime of high clouds in models and concluded that subsaturation expressed in terms of specific humidity is more useful than RH. At low temperatures, the low saturation vapor pressure limits how much condensate can evaporate regardless of the RH. This is consistent with Bardakov et al’s observation.

L287-288: “Interestingly, ...” Is this simply because the air density decreases with height so that the updrafts cannot support large hydrometeors?

L313-314: “Late cell phase samples... unavailable...” There’re no storms passing over T3 in Q4 of their lifecycle?