Ocean precursors to the extreme Atlantic 2017 hurricane season

Active Atlantic hurricane seasons are favoured by positive precursor sea surface temperature anomalies (SSTA) in the main development region (MDR, 10–20°N, 20–80°W). Here, we identify a different driving mechanism for these anomalies in 2017 (most costly season on record) compared to the recent active 2005 and 2010 seasons. In 2005 and 2010, a weakened Atlantic Meridional Overturning Circulation is the primary driver of positive SSTA. However, in 2017, reduced wind-driven cold water upwelling and weaker surface net heat loss in the north-eastern MDR were the main drivers. Our results are the first to show that air-sea heat flux and wind stress related processes are important in generating precursor positive SSTAs and that these processes were active pre-determinants of the 2017 season severity. In contrast to other strong seasons, positive SSTA developed later in 2017 (between April and July rather than March) compounding the challenge of predicting Atlantic hurricane season severity.

The manuscript by Hallam et al. presents an observational/reanalysis-based analysis comparing the atmospheric (primarily vertical wind shear) and oceanic factors that supported the active 2005, 2010, and 2017 Atlantic hurricane seasons. The main finding is that the 2017 season was unusual in that air-sea heat flux and wind-stress anomalies contributed to the SST warming favorable for tropical cyclones. This is an interesting and important topic, given the destructiveness of tropical cyclones and the relatively quick SST warming in 2017 that led to less skillful seasonal forecasts.
In my evaluation, the manuscript is publishable following major revisions. A better understanding of the oceanic factors that influence hurricane seasons is a valuable contribution, but I found that some parts of the analysis presented in the paper were unclear.
Major comments -One of the major findings of this study is that air-sea heat flux and wind-stress processes can be important driving factors for the warm SSTs supportive of active hurricane seasons. What processes usually drive such SSTAs? What other processes could be important? Some equations that outline the relevant terms for upper-ocean heat content would be helpful.
-section on Ocean heat flux time series: It is unclear why the discussion switches back and forth between LHFXA and MOC, and why the two are presented together in one (somewhat long) paragraph. Is there a connection between the two that is important to convey? -It is also unclear why the sections on "ocean and heat flux time series" and "hurricane season precursors" are separate, as the former discuses LHFX, which seems relevant to the SFXA discussion in the latter. It may be easier for a reader to understand if the sections are chosen according to ocean processes.
-The analysis on surface heat flux anomalies would also benefit from an equation that outlines the relevant terms. On line 243, the SFXA is attributed to specific components (which don't appear to be shown.). Similar comment for line 176, regarding the dominant contributions of LHFX.
-The results on SST and wind shear influences on hurricanes present some well-known relationships, which is fine in establishing that both atmospheric and oceanic factors supported the three active hurricane seasons. If space is tight, this analysis/discussion could be made more concise, in favor of a more thorough (perhaps equation-based) analysis of the oceanic processes that generated the SSTAs, which seems to be the more interesting and novel part of this study.
-lines 22-23: It is unclear how the two statements about " positive surface net heat flux anomaly" are different.
-line 64: 26.5C is outdated due to overall warming.
-lines 98 and 203: The mention of trends isn't really supported by an analysis and seems not central to the paper's focus. Unless there is a strong reason to bring up trends, I suggest reconsidering whether to discuss them, and adding more substance if choosing to keep them in the discussion.
-line 186: What is the interpretation for the different LHFXA in different regions? - Figure 1: Suggest modifying the yellow-red colorbar, which didn't print well in my copy.
Reviewer #3 (Remarks to the Author): This is an excellent work. The study used only observational (and reanalysis) data to show convincingly that three ocean variables (i.e., MDR surface latent heat flux, MOC at 26.5N and MDR wind stress curl) are important precursors to active Atlantic hurricane seasons such as 2005, 2010 and 2017. Unfortunately, MOC data is available only until Feb 2017. So, it is difficult to apply this hypothesis for the 2017 hurricane season. I encourage the authors to contact Rapid Mocha team to see if the MOC data for the later period can be used for this paper. Another related suggestion is to update the Ekman transport at 26.5N shown in Figure 3d to the end of 2017, which may tell us about the post Feb/2017 MOC values. Other than that, I only have some minor editorial comments.
Minor comments: 1) Line 83: "….leads a SSTA dipole in the North Atlantic" => It is not clear what "SSTA dipole in the North Atlantic" means. Describe it briefly here (e.g., where are the two poles located).
2) Lines 120-123: There is also a significant correlation between SSTA and shear anomaly of -0.58 (p<0.01) indicating that positive (negative) SSTA are often associated with negative (positive) shear anomalies, which has been linked to ENSO variability.

=>
It is not very clear what has been linked to ENSO variability. Is it SSTA, vertical wind shear or the link between SSTA and wind shear? Please revise the sentence to make it clear. SSTAs in MDR have been linked to ENSO, NAO and Atlantic meridional mode (AMM) (e.g., Enfield and Major, 1997;Czaja et al., 2002;Lee et al., 2008). MDR SSTAs -vertical wind shear link was shown in Wang and Lee (2007). ENSO-vertical wind shear link was shown in Goldenberg and Shapiro (1996)  Author's Response "Ocean precursors to the extreme Atlantic 2017 hurricane season", Hallam et al.
We are grateful to all three reviewers for their thoughtful and insightful comments that have enabled us to significantly improve the manuscript. A detailed response on each point is provided below.

Reviewers' comments:
Reviewer #1 (Remarks to the Author): Overall, this is a well-written and well-organized manuscript that provides a novel look at the precursors that generated the SSTA in the MDR region that led to the active 2017 Atlantic Hurricane season. The reviewer has raised an interesting point here. We have carried out analysis of the NCEP/NCAR reanalysis for 1980-2017 and find that the 2017 late spring/early summer heat flux anomalies are the most extreme positive values over the past nearly 40 years i.e. the weakest losses. We now provide more insight into how unusual these losses are through additional Supplementary Figure 3 and associated text in the main paper (section Ocean and heat flux time series).

Minor Comments:
L35-37: Is this for the North Atlantic region? Please specify. Yes, over 70% of total tropical cyclone damage in the North Atlantic is caused by major TCs, category 3, 4 or 5 on the Saffir-Simpson scale, which make landfall. The text has been updated.
L39-41: Again, the authors should mention that they are referring to the North Atlantic region. The text has been updated.

L61:
Year labels in figure 1b should be larger so that they are easier to read. The figure labels are now in a larger font.
L135-137: I suggest the authors mention/reference some of the important literature on rapid intensification, which seems critical to this point. For example, Lee et al., 2016Lee et al., (doi: 10.1038 provides significant evidence that the majority of major hurricanes undergo rapid intensification.
The text now includes mention of rapid intensification of hurricanes and references have been included. Figure 4b. The contour thickness has been increased for clarity.

L311
: Camp et al., 2018 suggested that they reasonably forecasted the spatial distribution anomaly of tropical cyclones with a May 15 seasonal forecast. Could you provide some insight, within your framework (i.e., did the model capture the SFXA?), as to why that might have been?
The reviewer raises an interesting point but to answer in detail regarding capturing the SFXA would require analysis of the Camp et al., model which we are not in a position to do. However, we do now note that Camp et al., find better representation of El Nino and this may have been a further factor in their reasonable forecast of the cyclone spatial distribution.
L336: Add the symbol for "delta T" after "ocean temperature anomalies". The text has been updated. Figure 2 and is referred to in the text in line 164.

Summary The manuscript by Hallam et al. presents an observational/reanalysis-based analysis comparing the atmospheric (primarily vertical wind shear) and oceanic factors that supported the active 2005, 2010, and 2017 Atlantic hurricane seasons. The main finding is that the 2017 season was unusual in that air-sea heat flux and wind-stress anomalies contributed to the SST warming favorable for tropical cyclones. This is an interesting and important topic, given the destructiveness of tropical cyclones and the relatively quick SST warming in 2017 that led to less skillful seasonal forecasts.
In my evaluation, the manuscript is publishable following major revisions. A better understanding of the oceanic factors that influence hurricane seasons is a valuable contribution, but I found that some parts of the analysis presented in the paper were unclear.

Major comments -One of the major findings of this study is that air-sea heat flux and wind-stress processes can be important driving factors for the warm SSTs supportive of active hurricane seasons. What processes usually drive such SSTAs? What other processes could be important? Some equations that outline the relevant terms for upper-ocean heat content would be helpful.
Variations in SST are influenced by the heat balance in the surface mixed layer, which comprises 3 main processes: surface fluxes, horizontal advection and vertical advection. The associated equation (1) is now included in the text with related discussion.
-section on Ocean heat flux time series: It is unclear why the discussion switches back and forth between LHFXA and MOC, and why the two are presented together in one (somewhat long) paragraph. Is there a connection between the two that is important to convey?
The addition of the equation for the heat balance in the surface mixed layer now makes it clearer why the two are presented together as they are the two key components influencing the upper ocean heat content and SSTA. The long paragraph has been split into two shorter ones for clarity.
-It is also unclear why the sections on "ocean and heat flux time series" and "hurricane season precursors" are separate, as the former discuses LHFX, which seems relevant to the SFXA discussion in the latter. It may be easier for a reader to understand if the sections are chosen according to ocean processes. The goal of the section "ocean and heat flux timeseries" is to provide an overview of the relation between ocean heat content/SSTs and the MOC and LHFX for all years. To us this seems an intuitive way of introducing the section on "hurricane season precursors", where we specifically zoom into the active seasons of 2005, 2010 and 2017 finding evidence of either ocean advection or LHFX being dominant. The section title has been updated accordingly.
We can see that the reviewer's suggestion could be a reasonable alternative method of presentation but we prefer to retain our present structure as it enables the advection and surface flux contributions to be considered together.
-The analysis on surface heat flux anomalies would also benefit from an equation that outlines the relevant terms. On line 243, the SFXA is attributed to specific components (which don't appear to be shown.). Similar comment for line 176, regarding the dominant contributions of LHFX.
The components of the net surface heat flux are now included within the description of the equation for the heat balance in the surface mixed layer and the relevant equation (2) is included in the methodology section.
-The results on SST and wind shear influences on hurricanes present some well-known relationships, which is fine in establishing that both atmospheric and oceanic factors supported the three active hurricane seasons. If space is tight, this analysis/discussion could be made more concise, in favor of a more thorough (perhaps equation-based) analysis of the oceanic processes that generated the SSTAs, which seems to be the more interesting and novel part of this study.
As noted above we have included a new equation (1) for the different processes involved to address the reviewer's concern on this point. Since space is not tight we prefer to retain the analysis / discussion as it is rather than cutting it back.
Minor comments -line 15: Suggest replacing "costliest ever" with "costliest on record". The manuscript has been updated.
-lines 22-23: It is unclear how the two statements about " positive surface net heat flux anomaly" are different. The negative wind stress curl anomaly and positive surface net heat flux anomaly occurred in April in the north eastern part of the MDR, whereas the positive surface flux anomalies between May and August 2017 occurred in the southern MDR region. This is now clarified in the text.
-lines 39 and 49: "active and intensive"seems redundant. The text has been updated.
-line 64: 26.5C is outdated due to overall warming. The text has been updated to include more recent studies and reference to the North Atlantic.
-lines 98 and 203: The mention of trends isn't really supported by an analysis and seems not central to the paper's focus. Unless there is a strong reason to bring up trends, I suggest reconsidering whether to discuss them, and adding more substance if choosing to keep them in the discussion. We understand the reviewer's concern and the wording has been amended accordingly.

-line 186: What is the interpretation for the different LHFXA in different regions?
The text has been updated to explain that the southern part of the MDR and north eastern MDR were chosen for further analysis because of the importance of these regions in 2017. Extreme LHFXA were seen in April in the north eastern MDR, and between May and August in the southern MDR.
- Figure 1: Suggest modifying the yellow-red colorbar, which didn't print well in my copy. The colorbar has been modified to remove the lighter shades of yellow.

Reviewer #3 (Remarks to the Author):
Review of "Ocean precursors to the extreme Atlantic 2017 hurricane season" by Samantha Hallam,Robert Marsh,Simon A. Josey,Joel Hirschi,Pat Hyder and Ben Moat This study investigated ocean precursors (i.e.,MDR surface latent heat flux,MOC at 26.5N and MDR wind stress curl) Figure 3d to the end of 2017, which may tell us about the post Feb/2017 MOC values. Other than that, I only have some minor editorial comments.
The RAPID team have confirmed to us that MOC data for 26.5N for 2017 will unfortunately not be available until mid 2019 as the data still needs to be collected from the RAPID array at 26.5N and then analysed. However, Ekman transport data at 26.5N is now shown in Figure 3d to the end of 2017. There was no decrease in the Ekman transport during 2017, and in view of the close correlation between the MOC and Ekman transports as seen in Figure 3d, supports the conclusion that the MOC did not play a role in the development of the positive SSTA in the MDR region during 2017. This is now reflected in the manuscript.
Minor comments: 1) Line 83: "….leads a SSTA dipole in the North Atlantic" => It is not clear what "SSTA dipole in the North Atlantic" means. Describe it briefly here (e.g., where are the two poles located). The text has been updated to reflect the location of the dipole pattern with poles 10-15N and 45-60N.