The authors leverage 15 years of GNSS radio occultation (RO) temperature profile data to examine the width of the tropics and its change over time. They compare the resulting diagnoses to longer records from several state-of-the-art reanalyses. Ultimately, it is demonstrated that RO data provide useful characterizations that broadly agree with the reanalysis diagnoses, especially upper troposphere lower stratosphere metrics where RO data are complete and most reliable. While I find the study to be mostly well-constructed and detailed, there are a few aspects that require a bit more clarification which I outline below.

General Comments

There are many instances of “on the NH” or “on the SH” that should all be revised to “in the NH” or “in the SH”.

Specific Comments

Line 7: the opening sentence of the abstract is a too strong of a statement. As the authors acknowledge later, this result is contingent upon the metric used. The language should be softened here.

Line 29: delete unnecessary period after “include”

Line 32: “systems” should be “system”

Lines 71-75: these sentences are entirely unnecessary

Line 85: a brief discussion of the wind retrieval is warranted here. It is later stated that the wind isn’t a simple geostrophic retrieval, so what is it?

Lines 92-93: is a ±2 day Gaussian time-weighting approach appropriate? This could be better justified/explained.

Line 250: “extend” should be “extent”

Line 257: “then” should be “than”

Line 311-312: there are a few studies that have revealed this narrowing via tropopause break metrics – Martin et al. 2020, https://doi.org/10.1175/JCLI-D-19-0629.1; Zou et al. 2023, https://doi.org/10.3389/feart.2023.1177502; Turhal et al. 2024, https://doi.org/10.5194/acp-24-13653-2024

Line 330: “this a globally” should be “this globally”

Line 335: delete unnecessary comma after “(Nimac et al., 2025a)”

This paper describes the use of GNSS-RO data to study several indirect metrics of the tropical width.  This includes the use of tropopause temperature (which has been done using RO before) and locations of the subtropical and eddy driven jets based on the newly derived winds fields obtained from the RO geopotential height data.  The authors show that the RO results generally compared well with the reanalyses, although some systematic (especially zonally varying) differences were noted.  The authors argued that the RO data period (2006 – 2020) was currently too small to detect the tropical width trends, with 35 years needed based on the reanalyses. 

This is a well-written paper with interesting results on an important topic.  I recommend its publication after the following comments are addressed.

Comments:

• In Sec 2.1, the authors should provide some description on the accuracy of the RO data used in the analysis, especially the wind retrieval which is a fairly new product that’s not known by most readers.
• It would be interesting to include the tropical width trends from the RO data in Fig 5. How does that compare with the reanalyses for the same 15 years?  How does that compare with previously published results if applicable (e.g, from Ao and Hajj 2012 , Davis and Birner 2013)
• Since the authors have 2D gridded dataset, can we say something about the regional trends? Is it possible that some regions show greater trends detectable over shorter periods?
• What can we say/do with the different (indirect) metrics giving different trends?
• Figure 4: some metrics show large zonal variation and systematic differences between RO and reanalyses over certain regions. Could you comment on implications on using the metric for monitoring tropical width changes, generally, or with RO?
• Abstract: “The tropics are expanding poleward as a result of anthropogenic climate change.” I think it’s more accurate to say that “Many studies have shown that the tropics are expanding…”