Dear Authors,

thank you very much for your dedication in answering almost all of the comments raised by the reviewers. Your paper has improved significantly because of this. I still have left some two minor, although important, issues to address:
* When you evaluate the GNSS-derived PWV with the PWV measured by radiosondes, you discuss the dry biases of the radiosondes, which is caused by solar heating (reference of Moradi et al. 2013). This dry bias is only try for certain types of radiosondes, so you should mention which type of radiosonde has been used at the three sites that you have used! Also the wet bias RS bias that you mention at high PWV ranges (reference of Bock et al. 2005) is radiosonde type dependent, see for instance a more receent publication of the same author: https://www.atmos-meas-tech.net/6/2777/2013/. An alternative explanation (a different sensitivity of GNSS and RS to different PWV ranges) of the GNSS-RS PWV differences is also given in e.g. https://www.atmos-meas-tech.net/7/2487/2014/. Of course, you can contact me for further details.
* when you evaluate the GNSS-derived PWV with the ERA-I PWV, you make the rather general statement that the performance of the high-resolution (O.25°X0.25°) ERA-I PWV product degrades with increased elevation. Please give some references here, as the ERA-I PWV product has been already evaluated frequently and you should find some evidence for this statement in such previous studies. A recent study (in which other references are found) is e.g. https://www.atmos-chem-phys-discuss.net/acp-2018-137/

And then, some technical corrections:
* page 1, line 30: water vapor accounts for only 0.1-3% of the total atmosphere --> in mass? density? Please specify.
* page 2, line 4: to estimated instead of of to estimating
* page 2, line 4: low-operating expenses instead of low-operating expense?
* page 2, line 20: add a ":" after needs two key meteorological parameters
* page2, line 21: add an "a" before barometer
* page 2, line 22: add "are" before not equipped
* page 5, line 10: equals to instead of equates to
* page 6, line 10: please add "adopted radius" in between This ensures...
* page 15, line 22: Referring to instead of Refer to
* page 17, line 8: I guess you want to add "eastern" to the regions where you find a high negative correlation between precipitation and PWV between 26°N and 27°N
* page 17, lines 8-9: please also refer here to your Figure 1, in which the elevation of the studied area is shown.
* page 18, Figure 11: please mention in the figure caption if the correlation coefficients are calculated for the accumulated precipitation and accumulated PWV for the 6-8 June 2015 period.

Dear authors,

just two small comments:
* on page 4, lines 11-12: please give a reference for the Chinese radiosonde types.
* on page 11, lines 15-16: please specify that the Chinese radiosonde types are known to have a moist bias for large PWV ranges (Nash et al. 2011 reference) and a similar effect has been found for other radiosonde types and was ascribed to the humidity sensors suffering from contamination from rain and clouds during radiosonde ascents (Bock et al. 2005 reference).

And last, but not least: I would encourage you to write an e-mail to the AMT office and ask if the paper could be included in the GNSS special issue https://www.atmos-meas-tech.net/special_issue89.html. This would increase the visibility of your paper and research considerably, as all the recent papers dealing with GNSS tropospheric research can be found at one place, after one click!

With best regards,

Roeland Van Malderen