Thank you for these very thoughtful comments, which we feel will improve the manuscript greatly.

The first major comment regarding the connection of convective MAPE to high percentile CAPE inspired an additional direction which we feel strengthens the paper. We have removed the discussion of high percentile CAPE, and we instead added discussion and analysis that compares convective MAPE with the occurrence of intense convective events using a precipitation feature database from the Global Precipitation Mission, where we found notable correspondence. Please see the updated Figure 6 and discussion of that figure in section 4.3.2 of the new manuscript.

In response to the second major comment regarding the skillful identification of the necessary conditions for WCBs, we have adjusted the statement in the abstract to read “This maximum potential ascent can be calculated based only on mean temperature and humidity, and WCB's tend to start in regions of high maximum potential ascent on a given day.” This is supported by  Figure 8 which demonstrates that on daily timescales WCB starting points occur in the areas of maximum potential ascent. The maxium potential ascent is not simply picking out storm track regions, as there is spatial heterogeneity on daily timescales within the storm track. Furthermore the spatial and distributions and seasonal cycles of climatological nonconvective MAPE and Eady growth (Fig 5) are different from those of maximum potential ascent and WCB starting points (Fig 7). For example, these distributions are quite different in the Southern Hemisphere in DJF or in the Northern Hemisphere in JJA as now pointed out in section 5.1.

In response to the third major comment regarding what MAPE adds over the Eady growth rate: while the Eady growth rate is easier to calculate, there is value in the unified theory that MAPE provides in this context, which includes energetics and ascent. As a thermodynamic measure of magnitude of eddies, non-convective MAPE has advantages over Eady growth rate in that it (1) includes the effects of latent heating whereas the Eady growth rate is a dry quantity applied to a moist atmosphere, and (2) it provides an energetic measure which can be directly compared to eddy kinetic energy. These two points are made in section 4.3.1. Furthermore, we emphasize that the Eady growth rate does not adequately capture the seasonality or spatial distributions of WCB starting points (e.g., compare the Southern Hemisphere in DJF in  figure 5 versus figure 7). This is now emphasized at the end of section 5.1.  

The reviewer asks if the ascent predicted by MAPE can be related to ascent in WCB trajectories. That is indeed the case: there is numerical consistency in the threshold of ascent used to define WCBs in the Madonna et al climatology (600hPa) and the values of maximum possible ascent for days with WCBs (see fig 9a; red curve) which have a median value of 570hPa as now emphasized in the text in section 5.2.

The reviewer asks for discussion of the large areas where there is ascent predicted by the MAPE calculation, but no WCBs. This is a helpful clarification – we consider the concept of maximum potential ascent similar to the idea of maximum potential intensity in tropical cyclones in the sense that the calculation reveals the thermodynamic potential for an event which may not necessarily be realized at all times. We have added language to this effect in section 5.2.

We have address both minor comments in the new manuscript as well. We have added a citation to better support the statement, “More than half of extratropical cyclones are associated with a WCB in northern hemisphere winter…”, and we have corrected the caption on Figure 2.

Thank you for these insightful comments, which have led to changes that we feel improve the manuscript.

Like Reviewer 1, Reviewer 2 asked about advantages of MAPE over the Eady growth rate. We highlight that the MAPE calculations presented in this paper offer a unified framework that connects to multiple weather aspects (please see reply on RC1 for further discussion). We add that the Eady growth rate is not an energetic measure, and it does not take latent heating into account, both of which are noted in section 4.3.1 of the manuscript. Also, the Eady growth rate does not provide information about occurrence of WCB starting points which have a distinct spatial pattern and seasonality compared to the Eady growth rate (compare figures 5 and 7, e.g. in the Southern Hemisphere in DJF) as now emphasized in the text in section 5.1. Lastly, the Eady growth rate does not give us information about convection, whereas we find that convective MAPE has some correspondence to regions of intense convection (see the revised figure 6).

We agree with the reviewer that it would be worthwhile to further develop the link from MAPE maximum potential ascent to WCBs, especially by developing a simple analytical expression for maximum potential ascent (akin to the Eady growth rate but predicting maximum potential ascent instead of growth). We have mentioned this possibility in the conclusions section, but it is non-trivial to develop such a theoretical expression and beyond the scope of the present work.

The reviewer suggested helpful ways to simplify the paper, and we have made changes that we feel improve the readability of the paper significantly. We have moved the discussion of algorithmic approaches to MAPE calculations from the introduction to the methods section 3 and shortened it. We have also added a sentence to direct readers who are only interested in the results before the description of the exact approach to calculating MAPE (just prior to section 3.1).

The reviewer also asks about why we chose to show both DJF and JJA in Figs. 3-7. We think it is important to show both DJF and JJA seasons because it helps to show correspondence between different aspects of MAPE and features such as WCBs. Furthermore,  it would be confusing to only show the results for one season for some aspects and both seasons for others. We added text to section 4.2 to point out that the locations of high local MAPE are different in summer versus winter due to the evolution of land-ocean contrasts in horizontal temperature gradients and static stability over the seasonal cycle.

The reviewer also asked about our choice of ERA-Interim as a reanalysis product. We chose ERA-Interim reanalysis as it was regarded as the best product at the time the analysis was done. The fact that we are relying on only monthly data (or daily data in one section) rather than subdaily data, and the agreement in major features in midlatitude energetics described in O’Gorman (2010), which uses NCEP2 reanalysis, and Gertler and O’Gorman (2019), which uses ERA-Interim reanalysis allows us to feel confident that our results are not product dependent.

We have addressed the typos identified by the reviewer in the new manuscript.