the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 05 Sep 2023
Effects of soil moisture and surface heat fluxes on the South American Monsoon System over West-Central Brazil: an observational study
Abstract. This study evaluated the relationship between different surface hydrometeorological variables and rainfall during the wet period of the South American Monsoon System (SAMS). A climatological study was performed for 30 rainy periods of the SAMS between the years 1991–2021 over the Central-West region of Brazil (WCB) (20–10° S and 60–50° W). The European Centre for Medium-Range Weather Forecasts Reanalysis 5th/ERA5 was used to understand, under different soil moist conditions (dry, intermediate, and wet), the hydrometeorological patterns and their impacts on SAMS during the three stages of the wet period: the development (September-October-November, SON), maturity (December-January-February, DJF), and weakening (March-April-May, MAM) South American monsoon quarters. The results show that along with an increase (decrease) in rainfall during the rainy season, there is also a significant increase (decrease) in both surface and subsurface volumetric soil moisture (θ) for the wet (dry) soil condition periods. However, the surface heat flux composites showed that the latent heat fluxes to the atmosphere (Hl) significantly exceeds the climatology during the SAMS development quarters (SON) for the wet soil group. In contrast, for the dry soil group, the significant increase of Hl, compared to the climatology, over the WCB occurred only during the SAMS maturity quarters (DJF), representing, in this last case, a significant injection of latent heat and consequently delayed evaporation (EV P) compared to rainy periods with wet soil, due to the low soil moisture content and the prevalence of dry convection over WCB. Regarding the sensible heat fluxes (Hs), it was observed drier (wetter) soils tend to exhibit values above (below) the climatological mean throughout all stages of SAMS evolution over the WCB. The 2-m air temperature (T 2m) and planetary boundary layer height (P BLH) anomaly composites showed that for the wet (dry) soil group and with significant rainfall above (below) the climatological mean over the WCB, all evolution quarters were marked by a significant decrease (increase) in T 2m and P BLH anomalies. Regarding the mechanisms of direct feedback between surface variables and rainfall associated with the SAMS, significant direct correlations were also observed between the mean rainfall of the SAMS rainy season and the mean values of active days duration (Dad), Hl, Hs and the surface Soil Moisture Condition Index (SMCI1).
- Preprint
(12589 KB) - Metadata XML
-
Supplement
(1865 KB) - BibTeX
- EndNote
Status: closed
-
AC1: 'Comment on hess-2023-198', João Pedro Gonçalves Nobre, 27 Sep 2023
Reviewing the manuscript, I noticed that some details could be improved:
- The mention of the Atlantic ITCZ and its connection to rainfall in the paragraph at line 146 could be removed. This is to avoid confusion for the reader about how this process connects to the recurrent rainfall throughout the entire wet season. Rephrasing this paragraph for a more general description would be a good alternative.
- The paragraphs describing surface soil moisture could be combined to make the text read more smoothly. The same is true for the paragraphs describing subsurface soil moisture.
- It would be helpful to standardize the term "stages" when referring to the evolution of the rainy season. At times throughout the text, "stages" is used, while at other times "phases" is used.
- In general, section 3.2.1 "Soil moisture and rainfall" could be revised in such a way that the reader can have a more fluid reading of the results. The suggestion would be to combine some paragraphs that deal with the same topic.
- At line 256, it would be helpful to change "moisture near the surface" to "at the surface".
- The paragraphs at lines 286 and 292 could be removed. The patterns for dry and intermediate soils could be described together, as they have similarities.
- A brief discussion could be added at the end of the section to orient the reader on the main findings of the section.
- The paragraphs at lines 328 and 338 need to be revised to improve comprehension of the described content.
- Throughout the text, the acronym SAMS is sometimes written as SMAS. It would be helpful to correct the acronym.
- The paragraph at line 468 needs to be revised, as the statement "soil moisture and surface evaporation decrease due to the reduction in evaporation and the associated latent heat flux" is redundant. The suggestion would be "during the demise of the SAMS, the precipitation, soil moisture and surface evaporation decrease, resulting in a higher proportion of thermal energy being transferred as sensible heat to the surface".
- As shown in the correlation matrix, the correlation between precipitation and Bowen ratio is negative. Therefore, lines 484-484 should be changed to "moisture content (SMCI1), but for Bowen ratio (Bo), a negative correlation with precipitation was verified".
- At line 502, it would be helpful to change "humidity values (not saturated environment with relative low actual vapor pressure compared with wet soils)" to "humidity values (with relative low actual vapor pressure compared with wet soils)" to avoid redundancy.
Citation: https://doi.org/10.5194/hess-2023-198-AC1 -
RC1: 'Comment on hess-2023-198', Anonymous Referee #1, 04 Oct 2023
*General Comments:
The authors proposed to study the coupling between soil moisture anomalies, surface heat fluxes and precipitation over a region deeply connected to the South American Monsoon System. The relevance of the topic and the motivation for their study is somewhat clear but could be improved. Unfortunately, the methods, results and discussions presented by the authors are either incomplete, unclear or not sufficient to prove their statements. The paper overall is poorly written, with a lot of missing information (like variable definitions or method descriptions) as well as grammatical errors, which makes it hard to understand what the authors have done. The authors make a lot of statements that are not backed up by any figure or analysis. Despite being a manuscript with a focus on precipitation, only one figure shows precipitation values. The arguments given by the authors to explain coupling between variables is very simplistic, and it is based on seasonal averages/anomalies and one correlation map. This is not sufficient to disentangle the complexity of soil moisture-precipitation (SM-P) coupling. A lot of methods can be found in the literature to analyze SM-P coupling, from coupling metrics to lagged correlations to model experiments with prescribed soil moisture, just to name a few. The authors should include more of these methods in their work to improve the robustness of their results. Furthermore, the authors do not discuss other sources of variability that influence the SAMS, like the ENSO. Unfortunately, I believe the manuscript needs major changes and it should not be considered for publication in its current form.
*Specific comments:
*The title says "an observational study" but the work only analyzes ERA5, which is a reanalysis and not an observational dataset. Please delete "an observational study" from the title.
*Line 3. Either change the acronym to CWB or change the name to West Central Brazil or similar, so the order of letters is consistent. Even more, in the paper title and in Figure 1 you call the region “West-Central Brazil”.
*Lines 20-21. At the end of the abstract, “active days duration” and “surface soil moisture condition index” are mentioned without previous introduction. I believe these concepts are not so generally known and you should not expect the reader to know them. I suggest either making a very short description of them or what they imply, talking in more general terms without referring to these specific terms.
*Lines 23-26. The way this sentence is written is confusing, because you point to different regions and different times for the beginning and demise of the SAMS. All regions where the SAMS is present will have an onset and demise period, it is not clear why you only mention some regions for onset and then different for demise. Please rephrase.
*Lines 27-29. Such a long list of references is not necessary here, you are not stating any specific point. Please cite the references more appropriately only when you need to back up your claims. For example, the sentence between lines 30-31 needs a reference (maybe is one of the references you stated previously, but you cannot expect the reader to check 10 references to find the right one).
*Lines 40-41. It is better to put the references after your statement, especially when you cite several references, so as to not interrupt the reading.
*Lines 55-56. It is not clear why the region WCB is mentioned here and why you chose it as a study case. Why is it important to focus on WCB? Throughout the Introduction you only talk about the SAMS in general.
*Lines 57-67. The period of study should be mentioned in this paragraph.
*Lines 70-75. I see now some explanation on why focus on the WCB region. This should be presented also in the Introduction, and it would be wise to add more background information on why this area is key for the development of the SAMS (with references). In the section Materials and Methods I suggest only presenting the limits of the region and the data.
*Section 2. There is no “Data” subsection in the “Materials and Methods” section. I think it is important to clearly present to the reader where the data information is. I would replace the “The region” subsection with a “Data” subsection and put there the info about ERA5 and the limits of the region. I would also state here the period considered.
*Lines 91-92. This statement needs some references.
*Lines 94. I believe it is enough to say that the Student t-test was used to test the statistical significance of the anomalies.
*Lines 100-101. I believe a short clarification of what are active and break periods could be useful here.
*Section 2.3 Clustering Algorithm. This section mostly presents the soil moisture variable and the SMCI, and the clustering method is barely mentioned. I understand that the k-means method is widely known, but then the section should not be called “Clustering Algorithm” if you are not going to talk about that. “Soil moisture conditions classification” sounds more appropriate.
*Line 113. You say “an average area over WCB between 20-10◦ S and 60-50◦ W” and to me it sounds like you are specifying a different region within WCB, and makes me go check again the region definition. Since WCB limits have already been defined, then just saying “the spatial average over WCB” is more straightforward.
*Lines 112-121. How do you account for seasonal variability? you should use daily anomalies and not daily averages. Are SSM max and min calculated on a per-calendar-day basis or do you consider single values for the whole period? That is, values from October should not be directly compared to values from January which are typically more humid.
*How do you know that the soil moisture-atmosphere coupling is strong or relevant in the region? At least a simple coupling metric should be calculated and shown.
*Lines 123-126. The method to identify the onset is not clear. Please rework this definition. Also, you say “In that study …” but it looks like you are describing your method and not the one from Gan et al (2004).
*Line 128. What do you mean by “we recalculated the daily average rainfall”? Is it a different calculation than the one in Section 2.2?
*Lines 126 and 133. It is not necessary to repeat the spatial definition of WCB.
*Lines 133-139. This Section needs to be rewritten. You mentioned twice that the MPI is the same as the one from Krishnamurthy and Shukla (2000) for India. But then you don’t explain how the index is constructed. You mention that the index is related to precipitation anomalies but then you compare the MPI to the standard deviation of the average precipitation, not anomalies.
*Lines 142-150. These lines are not describing results from this paper but from other papers, they should be moved to the Introduction. Here there should be a figure from your analyses showing the mean precipitation climatology in the different seasons.
*Lines 153-154. You mention here that “there is also an increase in the intensity of rainfall” but there is no figure showing rainfall.
*Figures 2 and 3. The colorbars in these figures are inappropriate. It is better to use a sequential colorbar from clearer to darker color or reversed, but not a divergent colorbar if you are not showing negative and positive values. A divergent colorbar here makes it difficult to correctly compare the subfigures.
*All Figures with map. I do not see the point in showing the whole of South America in the plots if only values inside WCB will be discussed.
*Lines 162-164. There is no analysis clearly showing the relation between Hs and T2m and atmospheric instability.
*Lines 162-165. Where do you get these temperature values? There is no Figure with temperature.
*Line 168-169. You cannot write these claims without showing results with temperature and precipitation.
*Line 171. “Therefore” means “as a consequence,”, but what you mention has nothing to do with what was said in the previous sentence.
*Lines 171-180. There are no figures of solar radiation, rainfall or atmospheric instability to back up all these claims.
*Lines 183-184. It is usually recommended to avoid these kinds of flamboyant claims and keep a modest tone.
*Lines 192-197. These lines are more suited for a conclusion. What is the point you are trying to make here?
*Figure 4 caption. Here it says that the composites were obtained from the monthly ERA5 dataset. But previously it was mentioned that the hourly ERA5 data was used (Line 83) and that daily averages and anomalies were calculated.
*The authors do not take any measure to consider the influence of ENSO, which is a main driver of SAMS variability.
*Lines 225-235. It is obviously expected that precipitation and soil moisture will be positively correlated, since precipitation anomalies will undoubtedly translate into soil moisture anomalies (unless focusing on an extremely dry or wet region). This correlation analysis does not constitute substantial proof that precipitation is modulated by soil moisture.
*Lines 237-245. There is no analysis in your work that supports these claims. Moisture recycling (moisture that enters the atmosphere from evaporation and then precipitates) is not the only mechanism that drives precipitation.
*Lines 246-251. This is, for example, a different mechanism. But why is it mentioned here? Do you encounter this mechanism in your study? You need to better link your ideas.
*Figure 7. Here you present correlations between rainfall and variables that were not introduced before, like Bo, and nothing is then analyzed with regards to those variables.
*Lines 261-267. There is no figure or analysis that provides evidence of these claims. You mention variables like cloud cover, atmospheric instability, atmospheric circulation and radiative forcing, but do not present any figure with those variables.
*Lines 326-333. This paragraph just describes results from previous works, and should be part of the Introduction. There is no clear connection between this part and what was described before.
*Line 366. These are important discussions that should have their own analysis, however the figure is not shown.
*Lines 341-375. A lot of information here was already said before. Either present all results and then make a general analysis, or construct the analysis along the presentation of results without repeating so much information.
*Lines 376-379. Here you analyze Bo, but this should have been said when the Figure 7 is presented.
*Variables Na, Dad, Nb, Dbd, Pi and Pf are sporadically mentioned but never introduced nor defined.
*Lines 388-393. The relationship that you are trying to show here is not clear to me.
*Lines 403-412. You did not show the climatological evolution of precipitation in your figures.
*A significant portion of the Conclusions is just repeating what was presented in the results, without a broader perspective.
*Technical corrections:
*Line 48. Citation not properly formatted.
*Line 85. “South American continent”
*Line 123. “we used a similar methodology as the one proposed by”
*Line 151. Why mention Fig. 3 before Fig. 2? Either flip the figures or fix the text.
*Line 309. Figure 11?
Citation: https://doi.org/10.5194/hess-2023-198-RC1 -
AC2: 'Reply on RC1', João Pedro Gonçalves Nobre, 05 Oct 2023
Dear Reviewer,
Thank you so much for all your suggestions. I am confident that they will help to significantly improve the quality of the manuscript.
Regarding the analyses that you question, I had attached them in the supplementary material. However, I agree that they should have been referenced in the main text so that the reader could understand what was being analyzed.
I also appreciate your willingness to read my manuscript. I will send you an update soon with all the changes I have made.
Best regards,
João.
Citation: https://doi.org/10.5194/hess-2023-198-AC2 -
AC4: 'Reply on RC1', João Pedro Gonçalves Nobre, 30 Oct 2023
Dear Reviewer,
I would like to once again thank you for all of your comments. They have greatly helped to improve the quality of the manuscript. I am very grateful for your dedication in providing such valuable feedback.
Here is a response to your specific comments:
*The title says "an observational study" but the work only analyzes ERA5, which is a reanalysis and not an observational dataset. Please delete "an observational study" from the title.
ANS: In the context in which I use the term observational study, it is to differentiate from an experimental study. Thus, the observational study indicated in this context that there is no manipulation of variables, we only observe and record events or behaviors as they occur. However, I agree that this part of the title can be removed to avoid confusion for the reader.
*Line 3. Either change the acronym to CWB or change the name to West Central Brazil or similar, so the order of letters is consistent. Even more, in the paper title and in Figure 1 you call the region “West-Central Brazil”.
ANS: Indeed, there are a number of times in the text where the abbreviation for West Central Brazil is spelled differently. I will only adopt WCB. Thank you very much for the observation.
Lines 20-21. At the end of the abstract, “active days duration” and “surface soil moisture condition index” are mentioned without previous introduction. I believe these concepts are not so generally known and you should not expect the reader to know them. I suggest either making a very short description of them or what they imply, talking in more general terms without referring to these specific terms.
ANS: In the next revision of the manuscript, we will add this information.
*Lines 23-26. The way this sentence is written is confusing, because you point to different regions and different times for the beginning and demise of the SAMS. All regions where the SAMS is present will have an onset and demise period, it is not clear why you only mention some regions for onset and then different for demise. Please rephrase.
ANS: This will be corrected in the new manuscript version.
*Lines 27-29. Such a long list of references is not necessary here, you are not stating any specific point. Please cite the references more appropriately only when you need to back up your claims. For example, the sentence between lines 30-31 needs a reference (maybe is one of the references you stated previously, but you cannot expect the reader to check 10 references to find the right one).
ANS: This will be corrected in the new manuscript version.
*Lines 40-41. It is better to put the references after your statement, especially when you cite several references, so as to not interrupt the reading.
ANS: I agree with this idea.
*Lines 55-56. It is not clear why the region WCB is mentioned here and why you chose it as a study case. Why is it important to focus on WCB? Throughout the Introduction you only talk about the SAMS in general.
ANS: In reality, I preferred to leave a more detailed explanation of the importance of WCB in the methodology section. However, in the original manuscript, I will make new adaptations so that this type of information is already included in the introduction.
*Lines 57-67. The period of study should be mentioned in this paragraph.
ANS: I agree with this point and the change will be made in a new version of the manuscript.
*Lines 70-75. I see now some explanation on why focus on the WCB region. This should be presented also in the Introduction, and it would be wise to add more background information on why this area is key for the development of the SAMS (with references). In the section Materials and Methods I suggest only presenting the limits of the region and the data.
ANS: Following the previous suggestion for a better description of the WCB, we will only leave the description of the region's boundaries in the materials and methods section.
*Section 2. There is no “Data” subsection in the “Materials and Methods” section. I think it is important to clearly present to the reader where the data information is. I would replace the “The region” subsection with a “Data” subsection and put there the info about ERA5 and the limits of the region. I would also state here the period considered.
ANS: We accept this suggestion. This change will be made in a new manuscript version.
*Lines 91-92. This statement needs some references.
ANS: References will be added in a new version of the manuscript.
*Lines 94. I believe it is enough to say that the Student t-test was used to test the statistical significance of the anomalies.
ANS: This change will be made in a new version of the manuscript.
*Lines 100-101. I believe a short clarification of what are active and break periods could be useful here.
ANS: This change will be made in a new version of the manuscript.
*Section 2.3 Clustering Algorithm. This section mostly presents the soil moisture variable and the SMCI, and the clustering method is barely mentioned. I understand that the k-means method is widely known, but then the section should not be called “Clustering Algorithm” if you are not going to talk about that. “Soil moisture conditions classification” sounds more appropriate.
ANS: I agree with this suggestion. In a new manuscript version, there will be a section as suggested.
*Line 113. You say “an average area over WCB between 20-10◦ S and 60-50◦ W” and to me it sounds like you are specifying a different region within WCB, and makes me go check again the region definition. Since WCB limits have already been defined, then just saying “the spatial average over WCB” is more straightforward.
ANS: The change will be made in a new version of the manuscript.
*Lines 112-121. How do you account for seasonal variability? you should use daily anomalies and not daily averages. Are SSM max and min calculated on a per-calendar-day basis or do you consider single values for the whole period? That is, values from October should not be directly compared to values from January which are typically more humid.
ANS: Note that the application of this normalization is obtained only for soil moisture data for days corresponding to the wet period of the rainy season of the South American Monsoon System. This methodology is not specifically applied to perform an intercomparison between the months/seasons belonging to the rainy season, but rather to obtain an average of the Soil Moisture Condition Index for each rainy season. The need for data normalization was interesting so that we could better capture the variability of the data when using the k-means clustering method. Therefore, we do not use this methodology specifically to capture seasonal variability, but to observe seasonal characteristics of different hydrometeorological variables when on average the soils throughout the wet period are dry, intermediate, or wet. Thus, when I cite "daily averages," I am referring to the daily average soil moisture data that I obtain from ERA5. In other words, it is as if I had a series of surface soil moisture data only for the days of the rainy season in a spreadsheet and then normalized them. We will keep this methodology. Given the type of question we want to answer, we believe that this methodology achieves our goal.
*How do you know that the soil moisture-atmosphere coupling is strong or relevant in the region? At least a simple coupling metric should be calculated and shown.
ANS: I agree that only presenting the quarterly compounds of different meteorological variables is not enough to prove the surface-atmosphere interactions. In relation to this point, we created some maps with some indices to show that these patterns do exist, but they are more significant in the development and demise phases of the SMAS rainy season. Please see the maps of the TCI, ACI, and TF indices for each soil type in the attached document. These are the same indices used by Chevuturi et al (2022) for South America (https://rmets.onlinelibrary.wiley.com/doi/epdf/10.1002/cli2.28).
*Lines 123-126. The method to identify the onset is not clear. Please rework this definition. Also, you say “In that study …” but it looks like you are describing your method and not the one from Gan et al (2004).
ANS: An amendment related to this will be made to the original text.
*Line 128. What do you mean by “we recalculated the daily average rainfall”? Is it a different calculation than the one in Section 2.2?
ANS: The intention was to say that we obtained a new climatological mean based on the ERA5 database for a new climatological period.
*Lines 126 and 133. It is not necessary to repeat the spatial definition of WCB.
ANS: Information will be removed.
*Lines 133-139. This Section needs to be rewritten. You mentioned twice that the MPI is the same as the one from Krishnamurthy and Shukla (2000) for India. But then you don’t explain how the index is constructed. You mention that the index is related to precipitation anomalies but then you compare the MPI to the standard deviation of the average precipitation, not anomalies.
ANS: This section will be rewritten as suggested.
*Lines 142-150. These lines are not describing results from this paper but from other papers, they should be moved to the Introduction. Here there should be a figure from your analyses showing the mean precipitation climatology in the different seasons.
ANS: In an earlier version, we had done this, but we removed it due to the large number of maps. In a new version, we decided to restructure the panels and insert them directly into the text, as per the panels that are included in the content attached to this message.
*Lines 153-154. You mention here that “there is also an increase in the intensity of rainfall” but there is no figure showing rainfall.
ANS: I finished inserting it in the supplementary material and not referencing it. In a new version of the manuscript, we will insert the climatological fields of rainfall as presented in the document attached to this message.
*Figures 2 and 3. The colorbars in these figures are inappropriate. It is better to use a sequential colorbar from clearer to darker color or reversed, but not a divergent colorbar if you are not showing negative and positive values. A divergent colorbar here makes it difficult to correctly compare the subfigures.
ANS: This change was made as recommended. Please see the new climatological panels created as suggested in the document attached to this message.
*All Figures with map. I do not see the point in showing the whole of South America in the plots if only values inside WCB will be discussed.
ANS: In fact, the initial discussions are focused on the WCB region, but further ahead there are discussions about significant patterns over areas further south in South America whose precipitation regime is similar to the COB. That is why I used a larger domain. In a new version of the article, I will reinforce the similarity of the patterns between these two regions throughout the text. In addition, although the focus of the study is on the WCB evaluation quadrant, sometimes immediately neighboring regions exhibit similar behavior, and if I focus the images only on the delineated COB region, I could lose some interesting details about surrounding areas.
*Lines 162-164. There is no analysis clearly showing the relation between Hs and T2m and atmospheric instability.
ANS: In a new version of the manuscript, this information will be better referenced.
*Lines 162-165. Where do you get these temperature values? There is no Figure with temperature.
ANS: In the supplementary material figures, Fig. S2. There is a missing reference to these figures in the text.
*Line 168-169. You cannot write these claims without showing results with temperature and precipitation.
ANS: These results are found in the supplementary material. Fig. S1 and S2.
*Line 171. “Therefore” means “as a consequence,”, but what you mention has nothing to do with what was said in the previous sentence.
ANS: I agree with the suggestion. In fact, since it refers to a summary of the analyzed content, it would be more convenient to use "In summary".
*Lines 171-180. There are no figures of solar radiation, rainfall or atmospheric instability to back up all these claims.
ANS: The output long-wave radiation field is found in the supplementary material. Figure S4. I will reference it in the text.
*Lines 183-184. It is usually recommended to avoid these kinds of flamboyant claims and keep a modest tone.
ANS: The changes will be made to the manuscript text in an attempt to maintain a more modest tone.
Lines 192-197. These lines are more suited for a conclusion. What is the point you are trying to make here?
ANS: In fact, we decided to remove the paragraph and keep only the most descriptive content of the results.
*Figure 4 caption. Here it says that the composites were obtained from the monthly ERA5 dataset. But previously it was mentioned that the hourly ERA5 data was used (Line 83) and that daily averages and anomalies were calculated.
ANS: In fact, we initially provided only a brief description of ERA5 data, mentioning that it is available hourly. However, in the work, we used both daily and monthly statistics. We will improve the description of the data used in the "Data" section to better inform the reader.
*The authors do not take any measure to consider the influence of ENSO, which is a main driver of SAMS variability.
ANS: In fact, the impacts of ENSO on central and southeastern Brazil are generally small (this has been reported in different studies). Significant impacts are only observed in areas of the North, Northeast, and South of Brazil, during the rainy season of the SAMS. In addition, previous analyses of different surface variables performed by me did not show significant relationships. Therefore, we decided not to focus our studies on this type of assessment, despite it being a great idea.
*Lines 225-235. It is obviously expected that precipitation and soil moisture will be positively correlated, since precipitation anomalies will undoubtedly translate into soil moisture anomalies (unless focusing on an extremely dry or wet region). This correlation analysis does not constitute substantial proof that precipitation is modulated by soil moisture.
ANS: TCI, ACI, and TF coupling indices will be used to support this statement, as shown in the supplementary material.
*Lines 237-245. There is no analysis in your work that supports these claims. Moisture recycling (moisture that enters the atmosphere from evaporation and then precipitates) is not the only mechanism that drives precipitation.
ANS: In fact, it requires some other fields to support this information. However, note that through the coupling indices (ACI, TCI, and TF) used to show how surface-atmosphere coupling mechanisms occur, attached in the supplementary material of this message, that during the development (SON) and withdrawal (MAM) quarters of the SAMS rainy season, there is generally a greater impact of surface variables on precipitation. However, in the maturity quarter (DJF), the presence of precipitation maxima is normally associated with a greater atmospheric control on surface variables, and this coincides with the period in which there is strong activity of the South Atlantic Convergence Zone over the region. In other words, the reported mechanism exists, but it needs to be better detailed in a new manuscript version.
*Lines 246-251. This is, for example, a different mechanism. But why is it mentioned here? Do you encounter this mechanism in your study? You need to better link your ideas.
ANS: Yes, this mechanism was verified in this work. And in a new version of the manuscript, the idea will be better linked throughout the sections.
*Figure 7. Here you present correlations between rainfall and variables that were not introduced before, like Bo, and nothing is then analyzed with regards to those variables
ANS: In a new version of this work, I will introduce this variable better.
*Lines 261-267. There is no figure or analysis that provides evidence of these claims. You mention variables like cloud cover, atmospheric instability, atmospheric circulation and radiative forcing, but do not present any figure with those variables.
ANS: These inferences were obtained from fields that I plotted but did not include in the manuscript, that is, by interpreting the compounds of 2-m air temperature anomaly, outgoing longwave radiation, and low-level geopotential anomaly. In a new review, I will add them.
*Lines 326-333. This paragraph just describes results from previous works, and should be part of the Introduction. There is no clear connection between this part and what was described before.
ANS: A new restructuring will be done, so that the ideas contained in this paragraph are better combined.
*Line 366. These are important discussions that should have their own analysis, however the figure is not shown.
ANS: The figure used to build this analysis was attached to the supplementary material, but it was not referenced. In a new version of the manuscript, we will add it to the main text.
*Lines 341-375. A lot of information here was already said before. Either present all results and then make a general analysis, or construct the analysis along the presentation of results without repeating so much information.
ANS: As suggested, elements throughout the paragraphs will be removed in order to avoid repetition.Lines 376-379. Here you analyze Bo, but this should have been said when the Figure 7 is presented.
ANS: An analysis will be added along with the latent and sensible heat fluxes at the surface.
*Variables Na, Dad, Nb, Dbd, Pi and Pf are sporadically mentioned but never introduced nor defined
ANS: An introduction to these variables will be added to the methodology.
*Lines 388-393. The relationship that you are trying to show here is not clear to me.
ANS: The superficial soil is more exposed to local weather conditions (temperature and evaporation, for example) and retains accumulated water for less time (which gives it greater variability within the wet period), compared to subsurface soils. Thus, since the beginning and end quarters of the rainy season are generally marked by a greater surface control on precipitation (see the attached figures for ACI, TCI, and TF), years in which subsurface soils are on average anomalously drier during the SAMS wet period tend to typically present a late start and early end of the SAMS wet period (this anomalous decrease only occurs when the surface layer suffers a water restriction). On the other hand, when the SAMS wet season, on average, has a relatively wet subsurface soil, there is a high chance of having an early start and late end of the SAMS wet season (this occurs when the surface soil is not suffering a water restriction).
*Lines 403-412. You did not show the climatological evolution of precipitation in your figures.
ANS: It is in the supplementary material, Figure S1. In the new version of the manuscript, I will better reference the figures.
*A significant portion of the Conclusions is just repeating what was presented in the results, without a broader perspective.
ANS: The conclusions will be rephrased to avoid repetition.
Technical corrections will be performed in the new version of the manuscript, as suggested.
Best regards,
João Nobre.
-
AC2: 'Reply on RC1', João Pedro Gonçalves Nobre, 05 Oct 2023
-
RC2: 'Comment on hess-2023-198', Anonymous Referee #2, 27 Oct 2023
This study analyzed the hydrometeorological characteristics during the South American Monsoon System (SAMS) over West-Central Brazil (WCB). At present, I think that the results of this study are insufficient to support the conclusions, and there is still a lack of analysis of the physical mechanisms of land-atmosphere coupling. I am not sure whether this work rises to the high level of quality and significant impact expected by HESS. Thus, I recommend rejecting this manuscript. My comments are provided below.
- Section 3.2.1: I think there may be a problem with the logic of this article. I do not agree that using composite analysis alone can prove that the soil moisture is a key variable in defining spatiotemporal precipitation patterns. On the contrary, perhaps the precipitation is an important factor in determining the spatiotemporal of soil moisture. Soil moisture may have an impact on precipitation through land-atmosphere coupling, but this study does not provide any sufficient evidence.
- The results section of this study cited many references to demonstrate the relevant results. What are your innovative points compared to these studies?
- Line 188: Only 6 years were identified as wet years in this study, and the composite analysis may have significant uncertainty due to the limited samples. ERA5 can provide long-term reanalysis data, and perhaps you can carry out the study in longer years.
- Lines 142-145、171-180: The spatiotemporal changes in precipitation and radiation should be provided to validate relevant statements.
- Line 237-241: The relationship between evapotranspiration and precipitation in this study is ambiguous. You think the increase in evapotranspiration is an important reason for the increase in precipitation, but there is a significant negative correlation between precipitation and evapotranspiration over WCB (Figure 7).
Minor Comments
- Table 1 is very non intuitive.
- Please indicate the supplement Figures in the manuscript.
- Figure 7 contains a lot of information, but not all of them are meaningful. Perhaps a more concise and clear expression can be used. In addition, what’s the mean of “0.1(.), 1()”? please check the figure caption.
- Figure 11: Please reverse the colorbar.
- Lines 349-351: please check this sentence.
- Line 357: “then” should be changed into “than”, please check it.
- Line 382: We generally do not use percentile form to represent the correlation coefficient.
- Please explain the meanings of Dad, Dbd, Na, Nb, Pi , Pf andSMCI4 when they first appear in the manuscript.
- Lines 439-442: Please check this sentence.
-
AC3: 'Reply on RC2', João Pedro Gonçalves Nobre, 27 Oct 2023
Dear Reviewer,
Thank you for your suggestions. Here is a response to your general comments:
- I agree that only presenting the quarterly compounds of different meteorological variables is not enough to prove the surface-atmosphere interactions. In relation to this point, we created some maps with some indices to show that these patterns do exist, but they are more significant in the development and demise phases of the SMAS rainy season. Please see the maps of the TCI, ACI, and TF indices for each soil type in the attached document. These are the same indices used by Chevuturi et al (2022) for South America (https://rmets.onlinelibrary.wiley.com/doi/epdf/10.1002/cli2.28).
- Please note that most of the studies cited focus on analyzing the patterns of coupling between the surface and atmosphere at beginning of the rainy season or on some of the months within it. A more in-depth analysis for the maturity (December-January-February) and intensification (March-April-May) quarters is not normally explored in the literature. In addition, this study also answers some questions that are not well explored in the literature, for example, is there any relationship between the number and duration of active and inactive days of the SAMS, as well as with the pentads of the beginning and end of the rainy season with the average rainfall over the Central-West region of Brazil during the SAMS wet period? According to the results, this relationship only exists between the average rainfall over Central-West Brazil and the duration of active days (the answer to the question asked is essential for different sectors of society, for example, agriculture and energy). Another discovery in the study, regarding anomaly compounds, is the behavior of latent heat fluxes to the surface in dry and wet soils. For wet soils, Significant positive anomalies are generally observed in the development quarter of the SMAS wet period. For dry soils, these positive anomalies are more easily verified in the following quarter (maturity). In other words, it is as if there is a delayed, abnormally significant injection of moisture into the atmosphere in the group of dry soils over West-Central Brazil. However, as shown by the ACI index (in the attached file), there is normally significant atmospheric control over surface processes in the maturity quarter, regardless of soil type. This may seem ambiguous with the statement made about the anomaly compounds of latent heat flux to the surface, but it is not. Although the maturity quarter of the SMAS rainy season for the soil group is controlled by the atmosphere over central areas of the Brazilian territory, this does not prevent the west-central region soil from injects water vapor into the atmosphere in the maturity quarter, during a period of rainfall inactivity, when the environment is less saturated.
- In relation to the amount of years used, we decided to use 30 rainy seasons of the South American monsoon system over the COB, because in our study we work with both daily and monthly ERA5 data. Note that, although monthly ERA5 data take up less space on the machine, high-resolution daily ERA5 data take up a lot of space, especially when considering the amount of variables used in some of the analyses in our study. In addition, focusing on the 30 rainy seasons between 1991-2021 was also an attempt to focus our study on a time series where we could obtain a more linear trend of our data (given the effects of climate change). Note that increasing the data period can catch biased data (to better understand this trend, the graph of TSM anomalies in the Nino 3.4 region and the CPC climatology used are available at https://origin.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ONI_change.shtml). In summary, the thirty years of data used is a good number for my analyses and as for the uncertainties, even if the amount of data is increased, they could remain.
- I agree with you about some of the information that is not included in the article. In this regard, a new panel structure was created to include climatological information that was mentioned in the text, but was not illustrated by images. In a first version, these images were included throughout the text, but we removed them in an attempt to reduce the number of images. We considered the possibility of removing the "Climatology" section, but we finally decided to keep it because it is interesting to have a reference to the climatology when we make climatological composites. A new version of the climatological maps is attached, if you are interested in viewing it.
- Indeed, the relationship between evapotranspiration and precipitation may seem ambiguous, but it is not. Note that the data in the correlation matrix evaluation refers to an average of daily data for the entire wet period. As mentioned earlier, in general, the relationship between surface heat fluxes and precipitation is more evident in the development and demise phases of the SMAS rainy season over central areas of the Brazilian territory. In the maturity quarter, there is a greater atmospheric control over the soil. However, when we average by the SMAS wet period, naturally rainy seasons marked by relatively drier soils will have more evaporation than rainy seasons marked by extremely wet soils, because relatively drier soils over central areas of the Brazilian territory tend on average to be associated with a less saturated environment, which allows for a greater injection of water vapor into the atmosphere, compared to wetter soils, which, due to their longer duration of active rainfall days (note this finding in the correlation matrix), contribute to a more saturated environment throughout the rainy season, which gives it lower evaporation rates (or lower latent heat fluxes to the surface).
Regarding minor reviews, we are working on each one of them.
In general, I agree that the work needs to be improved and that some major reviews need to be performed. However, I disagree with the point that the work should be directly rejected. As per the attached content, I can improve my discussions and provide more support for my discoveries.
I would like to thank you for your time and dedication in reading my work.
Best regards,
João Nobre.
Status: closed
-
AC1: 'Comment on hess-2023-198', João Pedro Gonçalves Nobre, 27 Sep 2023
Reviewing the manuscript, I noticed that some details could be improved:
- The mention of the Atlantic ITCZ and its connection to rainfall in the paragraph at line 146 could be removed. This is to avoid confusion for the reader about how this process connects to the recurrent rainfall throughout the entire wet season. Rephrasing this paragraph for a more general description would be a good alternative.
- The paragraphs describing surface soil moisture could be combined to make the text read more smoothly. The same is true for the paragraphs describing subsurface soil moisture.
- It would be helpful to standardize the term "stages" when referring to the evolution of the rainy season. At times throughout the text, "stages" is used, while at other times "phases" is used.
- In general, section 3.2.1 "Soil moisture and rainfall" could be revised in such a way that the reader can have a more fluid reading of the results. The suggestion would be to combine some paragraphs that deal with the same topic.
- At line 256, it would be helpful to change "moisture near the surface" to "at the surface".
- The paragraphs at lines 286 and 292 could be removed. The patterns for dry and intermediate soils could be described together, as they have similarities.
- A brief discussion could be added at the end of the section to orient the reader on the main findings of the section.
- The paragraphs at lines 328 and 338 need to be revised to improve comprehension of the described content.
- Throughout the text, the acronym SAMS is sometimes written as SMAS. It would be helpful to correct the acronym.
- The paragraph at line 468 needs to be revised, as the statement "soil moisture and surface evaporation decrease due to the reduction in evaporation and the associated latent heat flux" is redundant. The suggestion would be "during the demise of the SAMS, the precipitation, soil moisture and surface evaporation decrease, resulting in a higher proportion of thermal energy being transferred as sensible heat to the surface".
- As shown in the correlation matrix, the correlation between precipitation and Bowen ratio is negative. Therefore, lines 484-484 should be changed to "moisture content (SMCI1), but for Bowen ratio (Bo), a negative correlation with precipitation was verified".
- At line 502, it would be helpful to change "humidity values (not saturated environment with relative low actual vapor pressure compared with wet soils)" to "humidity values (with relative low actual vapor pressure compared with wet soils)" to avoid redundancy.
Citation: https://doi.org/10.5194/hess-2023-198-AC1 -
RC1: 'Comment on hess-2023-198', Anonymous Referee #1, 04 Oct 2023
*General Comments:
The authors proposed to study the coupling between soil moisture anomalies, surface heat fluxes and precipitation over a region deeply connected to the South American Monsoon System. The relevance of the topic and the motivation for their study is somewhat clear but could be improved. Unfortunately, the methods, results and discussions presented by the authors are either incomplete, unclear or not sufficient to prove their statements. The paper overall is poorly written, with a lot of missing information (like variable definitions or method descriptions) as well as grammatical errors, which makes it hard to understand what the authors have done. The authors make a lot of statements that are not backed up by any figure or analysis. Despite being a manuscript with a focus on precipitation, only one figure shows precipitation values. The arguments given by the authors to explain coupling between variables is very simplistic, and it is based on seasonal averages/anomalies and one correlation map. This is not sufficient to disentangle the complexity of soil moisture-precipitation (SM-P) coupling. A lot of methods can be found in the literature to analyze SM-P coupling, from coupling metrics to lagged correlations to model experiments with prescribed soil moisture, just to name a few. The authors should include more of these methods in their work to improve the robustness of their results. Furthermore, the authors do not discuss other sources of variability that influence the SAMS, like the ENSO. Unfortunately, I believe the manuscript needs major changes and it should not be considered for publication in its current form.
*Specific comments:
*The title says "an observational study" but the work only analyzes ERA5, which is a reanalysis and not an observational dataset. Please delete "an observational study" from the title.
*Line 3. Either change the acronym to CWB or change the name to West Central Brazil or similar, so the order of letters is consistent. Even more, in the paper title and in Figure 1 you call the region “West-Central Brazil”.
*Lines 20-21. At the end of the abstract, “active days duration” and “surface soil moisture condition index” are mentioned without previous introduction. I believe these concepts are not so generally known and you should not expect the reader to know them. I suggest either making a very short description of them or what they imply, talking in more general terms without referring to these specific terms.
*Lines 23-26. The way this sentence is written is confusing, because you point to different regions and different times for the beginning and demise of the SAMS. All regions where the SAMS is present will have an onset and demise period, it is not clear why you only mention some regions for onset and then different for demise. Please rephrase.
*Lines 27-29. Such a long list of references is not necessary here, you are not stating any specific point. Please cite the references more appropriately only when you need to back up your claims. For example, the sentence between lines 30-31 needs a reference (maybe is one of the references you stated previously, but you cannot expect the reader to check 10 references to find the right one).
*Lines 40-41. It is better to put the references after your statement, especially when you cite several references, so as to not interrupt the reading.
*Lines 55-56. It is not clear why the region WCB is mentioned here and why you chose it as a study case. Why is it important to focus on WCB? Throughout the Introduction you only talk about the SAMS in general.
*Lines 57-67. The period of study should be mentioned in this paragraph.
*Lines 70-75. I see now some explanation on why focus on the WCB region. This should be presented also in the Introduction, and it would be wise to add more background information on why this area is key for the development of the SAMS (with references). In the section Materials and Methods I suggest only presenting the limits of the region and the data.
*Section 2. There is no “Data” subsection in the “Materials and Methods” section. I think it is important to clearly present to the reader where the data information is. I would replace the “The region” subsection with a “Data” subsection and put there the info about ERA5 and the limits of the region. I would also state here the period considered.
*Lines 91-92. This statement needs some references.
*Lines 94. I believe it is enough to say that the Student t-test was used to test the statistical significance of the anomalies.
*Lines 100-101. I believe a short clarification of what are active and break periods could be useful here.
*Section 2.3 Clustering Algorithm. This section mostly presents the soil moisture variable and the SMCI, and the clustering method is barely mentioned. I understand that the k-means method is widely known, but then the section should not be called “Clustering Algorithm” if you are not going to talk about that. “Soil moisture conditions classification” sounds more appropriate.
*Line 113. You say “an average area over WCB between 20-10◦ S and 60-50◦ W” and to me it sounds like you are specifying a different region within WCB, and makes me go check again the region definition. Since WCB limits have already been defined, then just saying “the spatial average over WCB” is more straightforward.
*Lines 112-121. How do you account for seasonal variability? you should use daily anomalies and not daily averages. Are SSM max and min calculated on a per-calendar-day basis or do you consider single values for the whole period? That is, values from October should not be directly compared to values from January which are typically more humid.
*How do you know that the soil moisture-atmosphere coupling is strong or relevant in the region? At least a simple coupling metric should be calculated and shown.
*Lines 123-126. The method to identify the onset is not clear. Please rework this definition. Also, you say “In that study …” but it looks like you are describing your method and not the one from Gan et al (2004).
*Line 128. What do you mean by “we recalculated the daily average rainfall”? Is it a different calculation than the one in Section 2.2?
*Lines 126 and 133. It is not necessary to repeat the spatial definition of WCB.
*Lines 133-139. This Section needs to be rewritten. You mentioned twice that the MPI is the same as the one from Krishnamurthy and Shukla (2000) for India. But then you don’t explain how the index is constructed. You mention that the index is related to precipitation anomalies but then you compare the MPI to the standard deviation of the average precipitation, not anomalies.
*Lines 142-150. These lines are not describing results from this paper but from other papers, they should be moved to the Introduction. Here there should be a figure from your analyses showing the mean precipitation climatology in the different seasons.
*Lines 153-154. You mention here that “there is also an increase in the intensity of rainfall” but there is no figure showing rainfall.
*Figures 2 and 3. The colorbars in these figures are inappropriate. It is better to use a sequential colorbar from clearer to darker color or reversed, but not a divergent colorbar if you are not showing negative and positive values. A divergent colorbar here makes it difficult to correctly compare the subfigures.
*All Figures with map. I do not see the point in showing the whole of South America in the plots if only values inside WCB will be discussed.
*Lines 162-164. There is no analysis clearly showing the relation between Hs and T2m and atmospheric instability.
*Lines 162-165. Where do you get these temperature values? There is no Figure with temperature.
*Line 168-169. You cannot write these claims without showing results with temperature and precipitation.
*Line 171. “Therefore” means “as a consequence,”, but what you mention has nothing to do with what was said in the previous sentence.
*Lines 171-180. There are no figures of solar radiation, rainfall or atmospheric instability to back up all these claims.
*Lines 183-184. It is usually recommended to avoid these kinds of flamboyant claims and keep a modest tone.
*Lines 192-197. These lines are more suited for a conclusion. What is the point you are trying to make here?
*Figure 4 caption. Here it says that the composites were obtained from the monthly ERA5 dataset. But previously it was mentioned that the hourly ERA5 data was used (Line 83) and that daily averages and anomalies were calculated.
*The authors do not take any measure to consider the influence of ENSO, which is a main driver of SAMS variability.
*Lines 225-235. It is obviously expected that precipitation and soil moisture will be positively correlated, since precipitation anomalies will undoubtedly translate into soil moisture anomalies (unless focusing on an extremely dry or wet region). This correlation analysis does not constitute substantial proof that precipitation is modulated by soil moisture.
*Lines 237-245. There is no analysis in your work that supports these claims. Moisture recycling (moisture that enters the atmosphere from evaporation and then precipitates) is not the only mechanism that drives precipitation.
*Lines 246-251. This is, for example, a different mechanism. But why is it mentioned here? Do you encounter this mechanism in your study? You need to better link your ideas.
*Figure 7. Here you present correlations between rainfall and variables that were not introduced before, like Bo, and nothing is then analyzed with regards to those variables.
*Lines 261-267. There is no figure or analysis that provides evidence of these claims. You mention variables like cloud cover, atmospheric instability, atmospheric circulation and radiative forcing, but do not present any figure with those variables.
*Lines 326-333. This paragraph just describes results from previous works, and should be part of the Introduction. There is no clear connection between this part and what was described before.
*Line 366. These are important discussions that should have their own analysis, however the figure is not shown.
*Lines 341-375. A lot of information here was already said before. Either present all results and then make a general analysis, or construct the analysis along the presentation of results without repeating so much information.
*Lines 376-379. Here you analyze Bo, but this should have been said when the Figure 7 is presented.
*Variables Na, Dad, Nb, Dbd, Pi and Pf are sporadically mentioned but never introduced nor defined.
*Lines 388-393. The relationship that you are trying to show here is not clear to me.
*Lines 403-412. You did not show the climatological evolution of precipitation in your figures.
*A significant portion of the Conclusions is just repeating what was presented in the results, without a broader perspective.
*Technical corrections:
*Line 48. Citation not properly formatted.
*Line 85. “South American continent”
*Line 123. “we used a similar methodology as the one proposed by”
*Line 151. Why mention Fig. 3 before Fig. 2? Either flip the figures or fix the text.
*Line 309. Figure 11?
Citation: https://doi.org/10.5194/hess-2023-198-RC1 -
AC2: 'Reply on RC1', João Pedro Gonçalves Nobre, 05 Oct 2023
Dear Reviewer,
Thank you so much for all your suggestions. I am confident that they will help to significantly improve the quality of the manuscript.
Regarding the analyses that you question, I had attached them in the supplementary material. However, I agree that they should have been referenced in the main text so that the reader could understand what was being analyzed.
I also appreciate your willingness to read my manuscript. I will send you an update soon with all the changes I have made.
Best regards,
João.
Citation: https://doi.org/10.5194/hess-2023-198-AC2 -
AC4: 'Reply on RC1', João Pedro Gonçalves Nobre, 30 Oct 2023
Dear Reviewer,
I would like to once again thank you for all of your comments. They have greatly helped to improve the quality of the manuscript. I am very grateful for your dedication in providing such valuable feedback.
Here is a response to your specific comments:
*The title says "an observational study" but the work only analyzes ERA5, which is a reanalysis and not an observational dataset. Please delete "an observational study" from the title.
ANS: In the context in which I use the term observational study, it is to differentiate from an experimental study. Thus, the observational study indicated in this context that there is no manipulation of variables, we only observe and record events or behaviors as they occur. However, I agree that this part of the title can be removed to avoid confusion for the reader.
*Line 3. Either change the acronym to CWB or change the name to West Central Brazil or similar, so the order of letters is consistent. Even more, in the paper title and in Figure 1 you call the region “West-Central Brazil”.
ANS: Indeed, there are a number of times in the text where the abbreviation for West Central Brazil is spelled differently. I will only adopt WCB. Thank you very much for the observation.
Lines 20-21. At the end of the abstract, “active days duration” and “surface soil moisture condition index” are mentioned without previous introduction. I believe these concepts are not so generally known and you should not expect the reader to know them. I suggest either making a very short description of them or what they imply, talking in more general terms without referring to these specific terms.
ANS: In the next revision of the manuscript, we will add this information.
*Lines 23-26. The way this sentence is written is confusing, because you point to different regions and different times for the beginning and demise of the SAMS. All regions where the SAMS is present will have an onset and demise period, it is not clear why you only mention some regions for onset and then different for demise. Please rephrase.
ANS: This will be corrected in the new manuscript version.
*Lines 27-29. Such a long list of references is not necessary here, you are not stating any specific point. Please cite the references more appropriately only when you need to back up your claims. For example, the sentence between lines 30-31 needs a reference (maybe is one of the references you stated previously, but you cannot expect the reader to check 10 references to find the right one).
ANS: This will be corrected in the new manuscript version.
*Lines 40-41. It is better to put the references after your statement, especially when you cite several references, so as to not interrupt the reading.
ANS: I agree with this idea.
*Lines 55-56. It is not clear why the region WCB is mentioned here and why you chose it as a study case. Why is it important to focus on WCB? Throughout the Introduction you only talk about the SAMS in general.
ANS: In reality, I preferred to leave a more detailed explanation of the importance of WCB in the methodology section. However, in the original manuscript, I will make new adaptations so that this type of information is already included in the introduction.
*Lines 57-67. The period of study should be mentioned in this paragraph.
ANS: I agree with this point and the change will be made in a new version of the manuscript.
*Lines 70-75. I see now some explanation on why focus on the WCB region. This should be presented also in the Introduction, and it would be wise to add more background information on why this area is key for the development of the SAMS (with references). In the section Materials and Methods I suggest only presenting the limits of the region and the data.
ANS: Following the previous suggestion for a better description of the WCB, we will only leave the description of the region's boundaries in the materials and methods section.
*Section 2. There is no “Data” subsection in the “Materials and Methods” section. I think it is important to clearly present to the reader where the data information is. I would replace the “The region” subsection with a “Data” subsection and put there the info about ERA5 and the limits of the region. I would also state here the period considered.
ANS: We accept this suggestion. This change will be made in a new manuscript version.
*Lines 91-92. This statement needs some references.
ANS: References will be added in a new version of the manuscript.
*Lines 94. I believe it is enough to say that the Student t-test was used to test the statistical significance of the anomalies.
ANS: This change will be made in a new version of the manuscript.
*Lines 100-101. I believe a short clarification of what are active and break periods could be useful here.
ANS: This change will be made in a new version of the manuscript.
*Section 2.3 Clustering Algorithm. This section mostly presents the soil moisture variable and the SMCI, and the clustering method is barely mentioned. I understand that the k-means method is widely known, but then the section should not be called “Clustering Algorithm” if you are not going to talk about that. “Soil moisture conditions classification” sounds more appropriate.
ANS: I agree with this suggestion. In a new manuscript version, there will be a section as suggested.
*Line 113. You say “an average area over WCB between 20-10◦ S and 60-50◦ W” and to me it sounds like you are specifying a different region within WCB, and makes me go check again the region definition. Since WCB limits have already been defined, then just saying “the spatial average over WCB” is more straightforward.
ANS: The change will be made in a new version of the manuscript.
*Lines 112-121. How do you account for seasonal variability? you should use daily anomalies and not daily averages. Are SSM max and min calculated on a per-calendar-day basis or do you consider single values for the whole period? That is, values from October should not be directly compared to values from January which are typically more humid.
ANS: Note that the application of this normalization is obtained only for soil moisture data for days corresponding to the wet period of the rainy season of the South American Monsoon System. This methodology is not specifically applied to perform an intercomparison between the months/seasons belonging to the rainy season, but rather to obtain an average of the Soil Moisture Condition Index for each rainy season. The need for data normalization was interesting so that we could better capture the variability of the data when using the k-means clustering method. Therefore, we do not use this methodology specifically to capture seasonal variability, but to observe seasonal characteristics of different hydrometeorological variables when on average the soils throughout the wet period are dry, intermediate, or wet. Thus, when I cite "daily averages," I am referring to the daily average soil moisture data that I obtain from ERA5. In other words, it is as if I had a series of surface soil moisture data only for the days of the rainy season in a spreadsheet and then normalized them. We will keep this methodology. Given the type of question we want to answer, we believe that this methodology achieves our goal.
*How do you know that the soil moisture-atmosphere coupling is strong or relevant in the region? At least a simple coupling metric should be calculated and shown.
ANS: I agree that only presenting the quarterly compounds of different meteorological variables is not enough to prove the surface-atmosphere interactions. In relation to this point, we created some maps with some indices to show that these patterns do exist, but they are more significant in the development and demise phases of the SMAS rainy season. Please see the maps of the TCI, ACI, and TF indices for each soil type in the attached document. These are the same indices used by Chevuturi et al (2022) for South America (https://rmets.onlinelibrary.wiley.com/doi/epdf/10.1002/cli2.28).
*Lines 123-126. The method to identify the onset is not clear. Please rework this definition. Also, you say “In that study …” but it looks like you are describing your method and not the one from Gan et al (2004).
ANS: An amendment related to this will be made to the original text.
*Line 128. What do you mean by “we recalculated the daily average rainfall”? Is it a different calculation than the one in Section 2.2?
ANS: The intention was to say that we obtained a new climatological mean based on the ERA5 database for a new climatological period.
*Lines 126 and 133. It is not necessary to repeat the spatial definition of WCB.
ANS: Information will be removed.
*Lines 133-139. This Section needs to be rewritten. You mentioned twice that the MPI is the same as the one from Krishnamurthy and Shukla (2000) for India. But then you don’t explain how the index is constructed. You mention that the index is related to precipitation anomalies but then you compare the MPI to the standard deviation of the average precipitation, not anomalies.
ANS: This section will be rewritten as suggested.
*Lines 142-150. These lines are not describing results from this paper but from other papers, they should be moved to the Introduction. Here there should be a figure from your analyses showing the mean precipitation climatology in the different seasons.
ANS: In an earlier version, we had done this, but we removed it due to the large number of maps. In a new version, we decided to restructure the panels and insert them directly into the text, as per the panels that are included in the content attached to this message.
*Lines 153-154. You mention here that “there is also an increase in the intensity of rainfall” but there is no figure showing rainfall.
ANS: I finished inserting it in the supplementary material and not referencing it. In a new version of the manuscript, we will insert the climatological fields of rainfall as presented in the document attached to this message.
*Figures 2 and 3. The colorbars in these figures are inappropriate. It is better to use a sequential colorbar from clearer to darker color or reversed, but not a divergent colorbar if you are not showing negative and positive values. A divergent colorbar here makes it difficult to correctly compare the subfigures.
ANS: This change was made as recommended. Please see the new climatological panels created as suggested in the document attached to this message.
*All Figures with map. I do not see the point in showing the whole of South America in the plots if only values inside WCB will be discussed.
ANS: In fact, the initial discussions are focused on the WCB region, but further ahead there are discussions about significant patterns over areas further south in South America whose precipitation regime is similar to the COB. That is why I used a larger domain. In a new version of the article, I will reinforce the similarity of the patterns between these two regions throughout the text. In addition, although the focus of the study is on the WCB evaluation quadrant, sometimes immediately neighboring regions exhibit similar behavior, and if I focus the images only on the delineated COB region, I could lose some interesting details about surrounding areas.
*Lines 162-164. There is no analysis clearly showing the relation between Hs and T2m and atmospheric instability.
ANS: In a new version of the manuscript, this information will be better referenced.
*Lines 162-165. Where do you get these temperature values? There is no Figure with temperature.
ANS: In the supplementary material figures, Fig. S2. There is a missing reference to these figures in the text.
*Line 168-169. You cannot write these claims without showing results with temperature and precipitation.
ANS: These results are found in the supplementary material. Fig. S1 and S2.
*Line 171. “Therefore” means “as a consequence,”, but what you mention has nothing to do with what was said in the previous sentence.
ANS: I agree with the suggestion. In fact, since it refers to a summary of the analyzed content, it would be more convenient to use "In summary".
*Lines 171-180. There are no figures of solar radiation, rainfall or atmospheric instability to back up all these claims.
ANS: The output long-wave radiation field is found in the supplementary material. Figure S4. I will reference it in the text.
*Lines 183-184. It is usually recommended to avoid these kinds of flamboyant claims and keep a modest tone.
ANS: The changes will be made to the manuscript text in an attempt to maintain a more modest tone.
Lines 192-197. These lines are more suited for a conclusion. What is the point you are trying to make here?
ANS: In fact, we decided to remove the paragraph and keep only the most descriptive content of the results.
*Figure 4 caption. Here it says that the composites were obtained from the monthly ERA5 dataset. But previously it was mentioned that the hourly ERA5 data was used (Line 83) and that daily averages and anomalies were calculated.
ANS: In fact, we initially provided only a brief description of ERA5 data, mentioning that it is available hourly. However, in the work, we used both daily and monthly statistics. We will improve the description of the data used in the "Data" section to better inform the reader.
*The authors do not take any measure to consider the influence of ENSO, which is a main driver of SAMS variability.
ANS: In fact, the impacts of ENSO on central and southeastern Brazil are generally small (this has been reported in different studies). Significant impacts are only observed in areas of the North, Northeast, and South of Brazil, during the rainy season of the SAMS. In addition, previous analyses of different surface variables performed by me did not show significant relationships. Therefore, we decided not to focus our studies on this type of assessment, despite it being a great idea.
*Lines 225-235. It is obviously expected that precipitation and soil moisture will be positively correlated, since precipitation anomalies will undoubtedly translate into soil moisture anomalies (unless focusing on an extremely dry or wet region). This correlation analysis does not constitute substantial proof that precipitation is modulated by soil moisture.
ANS: TCI, ACI, and TF coupling indices will be used to support this statement, as shown in the supplementary material.
*Lines 237-245. There is no analysis in your work that supports these claims. Moisture recycling (moisture that enters the atmosphere from evaporation and then precipitates) is not the only mechanism that drives precipitation.
ANS: In fact, it requires some other fields to support this information. However, note that through the coupling indices (ACI, TCI, and TF) used to show how surface-atmosphere coupling mechanisms occur, attached in the supplementary material of this message, that during the development (SON) and withdrawal (MAM) quarters of the SAMS rainy season, there is generally a greater impact of surface variables on precipitation. However, in the maturity quarter (DJF), the presence of precipitation maxima is normally associated with a greater atmospheric control on surface variables, and this coincides with the period in which there is strong activity of the South Atlantic Convergence Zone over the region. In other words, the reported mechanism exists, but it needs to be better detailed in a new manuscript version.
*Lines 246-251. This is, for example, a different mechanism. But why is it mentioned here? Do you encounter this mechanism in your study? You need to better link your ideas.
ANS: Yes, this mechanism was verified in this work. And in a new version of the manuscript, the idea will be better linked throughout the sections.
*Figure 7. Here you present correlations between rainfall and variables that were not introduced before, like Bo, and nothing is then analyzed with regards to those variables
ANS: In a new version of this work, I will introduce this variable better.
*Lines 261-267. There is no figure or analysis that provides evidence of these claims. You mention variables like cloud cover, atmospheric instability, atmospheric circulation and radiative forcing, but do not present any figure with those variables.
ANS: These inferences were obtained from fields that I plotted but did not include in the manuscript, that is, by interpreting the compounds of 2-m air temperature anomaly, outgoing longwave radiation, and low-level geopotential anomaly. In a new review, I will add them.
*Lines 326-333. This paragraph just describes results from previous works, and should be part of the Introduction. There is no clear connection between this part and what was described before.
ANS: A new restructuring will be done, so that the ideas contained in this paragraph are better combined.
*Line 366. These are important discussions that should have their own analysis, however the figure is not shown.
ANS: The figure used to build this analysis was attached to the supplementary material, but it was not referenced. In a new version of the manuscript, we will add it to the main text.
*Lines 341-375. A lot of information here was already said before. Either present all results and then make a general analysis, or construct the analysis along the presentation of results without repeating so much information.
ANS: As suggested, elements throughout the paragraphs will be removed in order to avoid repetition.Lines 376-379. Here you analyze Bo, but this should have been said when the Figure 7 is presented.
ANS: An analysis will be added along with the latent and sensible heat fluxes at the surface.
*Variables Na, Dad, Nb, Dbd, Pi and Pf are sporadically mentioned but never introduced nor defined
ANS: An introduction to these variables will be added to the methodology.
*Lines 388-393. The relationship that you are trying to show here is not clear to me.
ANS: The superficial soil is more exposed to local weather conditions (temperature and evaporation, for example) and retains accumulated water for less time (which gives it greater variability within the wet period), compared to subsurface soils. Thus, since the beginning and end quarters of the rainy season are generally marked by a greater surface control on precipitation (see the attached figures for ACI, TCI, and TF), years in which subsurface soils are on average anomalously drier during the SAMS wet period tend to typically present a late start and early end of the SAMS wet period (this anomalous decrease only occurs when the surface layer suffers a water restriction). On the other hand, when the SAMS wet season, on average, has a relatively wet subsurface soil, there is a high chance of having an early start and late end of the SAMS wet season (this occurs when the surface soil is not suffering a water restriction).
*Lines 403-412. You did not show the climatological evolution of precipitation in your figures.
ANS: It is in the supplementary material, Figure S1. In the new version of the manuscript, I will better reference the figures.
*A significant portion of the Conclusions is just repeating what was presented in the results, without a broader perspective.
ANS: The conclusions will be rephrased to avoid repetition.
Technical corrections will be performed in the new version of the manuscript, as suggested.
Best regards,
João Nobre.
-
AC2: 'Reply on RC1', João Pedro Gonçalves Nobre, 05 Oct 2023
-
RC2: 'Comment on hess-2023-198', Anonymous Referee #2, 27 Oct 2023
This study analyzed the hydrometeorological characteristics during the South American Monsoon System (SAMS) over West-Central Brazil (WCB). At present, I think that the results of this study are insufficient to support the conclusions, and there is still a lack of analysis of the physical mechanisms of land-atmosphere coupling. I am not sure whether this work rises to the high level of quality and significant impact expected by HESS. Thus, I recommend rejecting this manuscript. My comments are provided below.
- Section 3.2.1: I think there may be a problem with the logic of this article. I do not agree that using composite analysis alone can prove that the soil moisture is a key variable in defining spatiotemporal precipitation patterns. On the contrary, perhaps the precipitation is an important factor in determining the spatiotemporal of soil moisture. Soil moisture may have an impact on precipitation through land-atmosphere coupling, but this study does not provide any sufficient evidence.
- The results section of this study cited many references to demonstrate the relevant results. What are your innovative points compared to these studies?
- Line 188: Only 6 years were identified as wet years in this study, and the composite analysis may have significant uncertainty due to the limited samples. ERA5 can provide long-term reanalysis data, and perhaps you can carry out the study in longer years.
- Lines 142-145、171-180: The spatiotemporal changes in precipitation and radiation should be provided to validate relevant statements.
- Line 237-241: The relationship between evapotranspiration and precipitation in this study is ambiguous. You think the increase in evapotranspiration is an important reason for the increase in precipitation, but there is a significant negative correlation between precipitation and evapotranspiration over WCB (Figure 7).
Minor Comments
- Table 1 is very non intuitive.
- Please indicate the supplement Figures in the manuscript.
- Figure 7 contains a lot of information, but not all of them are meaningful. Perhaps a more concise and clear expression can be used. In addition, what’s the mean of “0.1(.), 1()”? please check the figure caption.
- Figure 11: Please reverse the colorbar.
- Lines 349-351: please check this sentence.
- Line 357: “then” should be changed into “than”, please check it.
- Line 382: We generally do not use percentile form to represent the correlation coefficient.
- Please explain the meanings of Dad, Dbd, Na, Nb, Pi , Pf andSMCI4 when they first appear in the manuscript.
- Lines 439-442: Please check this sentence.
-
AC3: 'Reply on RC2', João Pedro Gonçalves Nobre, 27 Oct 2023
Dear Reviewer,
Thank you for your suggestions. Here is a response to your general comments:
- I agree that only presenting the quarterly compounds of different meteorological variables is not enough to prove the surface-atmosphere interactions. In relation to this point, we created some maps with some indices to show that these patterns do exist, but they are more significant in the development and demise phases of the SMAS rainy season. Please see the maps of the TCI, ACI, and TF indices for each soil type in the attached document. These are the same indices used by Chevuturi et al (2022) for South America (https://rmets.onlinelibrary.wiley.com/doi/epdf/10.1002/cli2.28).
- Please note that most of the studies cited focus on analyzing the patterns of coupling between the surface and atmosphere at beginning of the rainy season or on some of the months within it. A more in-depth analysis for the maturity (December-January-February) and intensification (March-April-May) quarters is not normally explored in the literature. In addition, this study also answers some questions that are not well explored in the literature, for example, is there any relationship between the number and duration of active and inactive days of the SAMS, as well as with the pentads of the beginning and end of the rainy season with the average rainfall over the Central-West region of Brazil during the SAMS wet period? According to the results, this relationship only exists between the average rainfall over Central-West Brazil and the duration of active days (the answer to the question asked is essential for different sectors of society, for example, agriculture and energy). Another discovery in the study, regarding anomaly compounds, is the behavior of latent heat fluxes to the surface in dry and wet soils. For wet soils, Significant positive anomalies are generally observed in the development quarter of the SMAS wet period. For dry soils, these positive anomalies are more easily verified in the following quarter (maturity). In other words, it is as if there is a delayed, abnormally significant injection of moisture into the atmosphere in the group of dry soils over West-Central Brazil. However, as shown by the ACI index (in the attached file), there is normally significant atmospheric control over surface processes in the maturity quarter, regardless of soil type. This may seem ambiguous with the statement made about the anomaly compounds of latent heat flux to the surface, but it is not. Although the maturity quarter of the SMAS rainy season for the soil group is controlled by the atmosphere over central areas of the Brazilian territory, this does not prevent the west-central region soil from injects water vapor into the atmosphere in the maturity quarter, during a period of rainfall inactivity, when the environment is less saturated.
- In relation to the amount of years used, we decided to use 30 rainy seasons of the South American monsoon system over the COB, because in our study we work with both daily and monthly ERA5 data. Note that, although monthly ERA5 data take up less space on the machine, high-resolution daily ERA5 data take up a lot of space, especially when considering the amount of variables used in some of the analyses in our study. In addition, focusing on the 30 rainy seasons between 1991-2021 was also an attempt to focus our study on a time series where we could obtain a more linear trend of our data (given the effects of climate change). Note that increasing the data period can catch biased data (to better understand this trend, the graph of TSM anomalies in the Nino 3.4 region and the CPC climatology used are available at https://origin.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ONI_change.shtml). In summary, the thirty years of data used is a good number for my analyses and as for the uncertainties, even if the amount of data is increased, they could remain.
- I agree with you about some of the information that is not included in the article. In this regard, a new panel structure was created to include climatological information that was mentioned in the text, but was not illustrated by images. In a first version, these images were included throughout the text, but we removed them in an attempt to reduce the number of images. We considered the possibility of removing the "Climatology" section, but we finally decided to keep it because it is interesting to have a reference to the climatology when we make climatological composites. A new version of the climatological maps is attached, if you are interested in viewing it.
- Indeed, the relationship between evapotranspiration and precipitation may seem ambiguous, but it is not. Note that the data in the correlation matrix evaluation refers to an average of daily data for the entire wet period. As mentioned earlier, in general, the relationship between surface heat fluxes and precipitation is more evident in the development and demise phases of the SMAS rainy season over central areas of the Brazilian territory. In the maturity quarter, there is a greater atmospheric control over the soil. However, when we average by the SMAS wet period, naturally rainy seasons marked by relatively drier soils will have more evaporation than rainy seasons marked by extremely wet soils, because relatively drier soils over central areas of the Brazilian territory tend on average to be associated with a less saturated environment, which allows for a greater injection of water vapor into the atmosphere, compared to wetter soils, which, due to their longer duration of active rainfall days (note this finding in the correlation matrix), contribute to a more saturated environment throughout the rainy season, which gives it lower evaporation rates (or lower latent heat fluxes to the surface).
Regarding minor reviews, we are working on each one of them.
In general, I agree that the work needs to be improved and that some major reviews need to be performed. However, I disagree with the point that the work should be directly rejected. As per the attached content, I can improve my discussions and provide more support for my discoveries.
I would like to thank you for your time and dedication in reading my work.
Best regards,
João Nobre.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 373 | 82 | 23 | 478 | 36 | 12 | 9 |
- HTML: 373
- PDF: 82
- XML: 23
- Total: 478
- Supplement: 36
- BibTeX: 12
- EndNote: 9
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1