Manuscript title: A rise in subweekly temperature variability over Southern Hemisphere landmasses detected in multiple reanalyses

Authors: Martineau et al.

The article explores the agreement between datasets in depicting subweekly variability in the Southern Hemisphere to then use the reanalysis datasets to explore the trend of this variability. Overall there is a good agreement between reanalysis, specially between the new generation (ERA5, MERRA2, etc). The authors find some positive trends in subweekly temperature variability although it depends on the season considered and the region. In addition, this trend in not always significant.The topic fits the scope of this journal and the manuscript is generally well-written. Therefore, the manuscript is valuable to be published after some changes.

Major comment: I understand that nowadays research articles have titles that are eye-catching and also highlights the main results, instead of summarizing the type of research conducted. However, from the analysis of Figure 5 I don’t think that the title corresponds well with the results. The authors find some positive significant trends in some regions of the SH landmasses (mainly midlatitudes) but this do not occur in all seasons (actually it is only in midlatitude South America in DJF and MAM and South Africa in SON and southern Australia in JJA). So I would like to ask the authors to change the title accordingly. Actually, considering that this article is part of a special issue on reanalysis I find more interesting the good agreement between reanalysis over the Southern Hemisphere in representing trends.

Methods:

I’m assuming that the authors use daily reanalysis data. The different reanalysis used in this study also have different temporal resolution (some of them has 6-hourly data, hourly data available, etc). However I don’t find in the document any reference to this. Could the authors briefly explain how they treat the differences in the temporal resolution between reanalysis as they did with the spatial resolution? Also a brief discussion on how these differences may impact the results would be appreciated.

Results:

Line 21: Talking about extratropical variability and immediately after mentioning tropical storms does not sound coherent to me. I’d use mesoscale storms or something similar

134: I’d remove the expression between commas “like the Antarctic polar frontal zone” as the Antarctic Polar frontal zone does not entirely owe its existence to L-S contrasts. Removing this won’t change the meaning of the sentence.

148-149: The relationship between the local maxima and stationary waves in the SH needs a reference. Otherwise, it should be removed. To my knowledge the wave 1 is the QS wave that dominates the variability in the Southern Hemisphere (see for instance Quintanar and Mechoso 1995).

Figure 3: I find the selection of SON a bit arbitrary. Can the authors briefly discuss other seasons and include the corresponding figures as supplementary material? Another option could be showing the biases only for reanalysis included in the REM in the main document, and put the remaining reanalysis in the supplementary material. There is also a label “cti: 1.00+e00” next to the bar that does not make any sense to me

Line 223: What do you mean by “clearer”? Are the trends higher or lower?

Line 304-306: I don’t understand how do you correlate the Tvar trend of each reanalysis (one value per reanalysis) against the reference

Conclusions:

Line 356: Chemke et al 2022 only refers to CMIP6 data. Are the authors sure that the comparison with CMIP5 data cited in the article comes from Chemke 2022? I could not find it. Nevertheless, I don’t find the reference to CMIP6/5 data useful at all since Chemke already pointed out that models do not represent the observed trend well and therefore might underestimate the future trend. I’d remove the reference to CMIP data there. The authors can still speculate on the agreement between future changes in EKE and low level temperature variability.

Line 345-346: I would not compare directly reanalysis data at 850hPa to surface station based data. It could be that all reanalysis have problems representing the trends at 850hPa and the conclusions of sfc processes amplifying the trends can’t be drawn from the comparison you just made (I agree with you that sfc processes may amplify trends but I can’t arrive to this conclusion from the results you have shown)

In this study, the authors assess the climatology and interannual variability of subweekly temperature variance among multiple reanalyses. Results show that there is a good agreement for the climatological temperature variance and dominant sources and sinks of variance. The authors also point out that there is a good agreement for the positive trends in subweekly variability over South Africa and South America. The analyses are clear, the results are reliable, and the writing is good. I have some concerns and suggestions for the authors to consider for improving their manuscript.

Comments:

1. What are the reasons for the largest bias of NCEP-NCAR (R1) and NCEP-DOE (R2)that are modern full-input datasets from the REM climatology? How the observarions of R1 and R2 assimilation are distributed? And how the assimilated obsevations can affect the subweekly variability and generation term F_horiz?
2. The Eq. (1) explains the tendency of T_VAR, not the climatology or trend of T_VAR. For equilibrium, the tendency of T_VARis close to 0 and F_horiz and F_vert always cancel each other out. The characteristics of climatolgical subweekly variability may not simply be explained by the source or sink terms the right-hand side of Eq. (1). Please integrate the right-hand side to obtain the source for T_VAR or differentiate the left side to obtain the tendency of T_VAR.
3. What is the time-scale of the trends of T_VARinvestigated in the manuscript? Whether the effects of global warming are included? What internal or external forcings can explain the trends. That is how can we understand the trend of efficiency term?
4. The manuscript focuses on the climatology and trend of different seasons of three landmasses in the Southern Hemisphere, which are mainly descriptive. Please give more conclusions about seasonal differences from the perspective of physical mechanisms, which may be more impressive.