This is a very thorough study of a relevant topic, with a number of interesting results.  The analysis is careful and detailed, and the manuscript is well organized and well written.  For these reasons I recommend publication with minor comments.

Here are those minor comments:

The use of CIWV to represent water vapor amount above a certain layer is non-standard and thus a bit confusing.  ‘Column integrated’ implies total in the entire column; a web search shows many examples of this usage:  https://www.google.com/search?client=firefox-b-1-d&q=column+integrated+water+vapor.  One suggestion is to use “Layer Integrated Water Vapor”, whose meaning should be immediately apparent.

A few more details on the analysis are needed in the Abstract. Please explicitly mention the A-Train instruments used and what quantities are examined.  Also, please expand briefly on the synergistic retrieval.  What specific instruments does it use?  A sentence or two should suffice.  Also, how are cooling rates estimated.  Again, a sentence or two will suffice.  Finally, change ‘convections’ to ‘convection’.

Line 118.  Please state which two (or more?) AIRS L1B frequencies are used to estimate brightness temperature.

Include the OT, NOT, etc. definitions in figure 2 caption.  Otherwise, terms for quantities in the figure cannot be fully understood without knowing where in the text the figure is discussed.

Line 339.  Change ‘cools’ to ‘cool’.

Line 463.  Change “under overcast cloud conditions” to “for overcast cloud conditions” or something similar.  Please don’t use “under” because it implies beneath the clouds.

General: The authors have chosen a topic, modification of the thermodynamic structure of the tropopause by organized convective systems, that is both timely and interesting. This is an important topic because an accurate understanding of diabatic convective processes at the tropopause in the tropics and subtropics could impact the modeling of climate response to increased convection caused by rising surface temperatures. The analysis uses retrievals of temperature and water vapor profiles from AIRS, combined with a Radar-Lidar estimation of cloud ice water content to address an old and on-going debate about whether overshooting convective plumes can hydrate the lowermost stratosphere.  It is refreshing that the authors have provided an alternative to the very coarse-resolution MLS water vapor profiles to address this question, with the strong benefit of having a co-located retrieved temperature profile. The cyclone-centered coordinates are a welcome way to organize the observations.  I am recommending that this paper be accepted with some minor changes that are listed in the attached document.  In addition, vertical resolution around the tropopause for each data set/retrieval needs to be stated explicitly, some references ought to be consulted and updated to reflect the most current thinking and the development/uncertainty of satellite algorithms, for DARDAR-cloud and for the brightness temperatures.  Uncertainties and potential sampling biases need to be discussed. Are the retrievals representative?  How many AIRS-DARDAR combined profiles did not converge (I am only finding the number that did converge).  What do the authors think about diurnal changes?  Are these relevant to their results?  The choice of 16 km as a threshold for "overshoots" seems arbitrary and creates awkwardness for the interpretation because it often includes the cold point.  Many tropical cyclones extend much higher than this at their cores, sometimes to 18 km. I'm not recommending that the authors redo the analysis, but acknowledgement that "overshoots" may instead be "cloud tops" ought to be included.  Presentation of the new AIRS retrieval technique adds information and shows good promise, and it would be good to also see the authors present and understand the limitations of this technique.