The manuscript has been revised thoroughly according to the reviewers’ comments. In the present form, this paper provides a comprehensible overview of post-processing routines run operationally at DWD, which is definitely relevant to the audience of this NPG special issue. Hence, I think the paper will be ready for publication subject to a very minor revision addressing the following issues:

- L50: EMOS does not necessary rely on Gaussian distributions (e.g. EMOS for precipitation based on shifted GEV or shifted Gamma distributions). So I would either call the method NGR (non-homogenous Gaussian regression) here, or rephrase the sentence in order emphasize that depending on the variable of interest EMOS methods relying on different parametric distribution families exist.
- Equation (1): Why is a subscript k attached to predictand y. This notation is somewhat unusual and needs to be clarified.
- L193: is this still the case (considering the developments in statistical post-processing over the last decade)?
- L213: What do you mean by variable here? NWP output?

Some typos (list probably not exhaustive):
- L48: forecast errors
- L63/64: use of calibrated ensembles
- L89: thunderstorms?
- L137: Too many observations
- L157: Bougeault et al. (2009); Swinbank et al. (2016)
- L479 : ranked probability score

Dear Reinhold,

I'm happy to inform you that the 2nd reviewer now supports publication as well. However, I've still spotted a few typos and sentences that are not so easy to understand. You can find a non-exhaustive list below. Therefore, I would really appreciate if you could carefully correct typos and sentences that are still a bit difficult to understand.

Best regards,
Stephan

--------------------------------

- Line 13: Ensemble-MOS
- Line 46: to the costs of --> at the cost of
- Line 55: What do you mean by ‘Many special systems’? Maybe something like ‘Many different post-processing methods tailored to different variables exist, only…’ would be easier to understand?
- Line 62/63: Difficult to understand. I would replace it by ‘E.g., Schefzik et al. (2013); Schefzik and Möller (2018) use ensemble copula coupling (ECC) and Schaake shuffle-based approaches in order to generate post-processed forecast scenarios for temperature, precipitation and wind.’
- L105: I would write something like ‘The results shown here focus on …’
- L201 and 202: Do you really mean ‘estimation’? I would call it an ‘estimate’.
- L355: Maybe ‘writing. Hence, the time series are subject to a number of model changes.’ instead of ‘writing, while a number of model changes have taken place.’ would be easier to understand
- L414: ‘TIGGE, see 2.3.’  ‘TIGGE, see Sec. 2.3.’
- Figure 10: Why are the consistency bars centered around the diagonal? Should the consistency bars not by more or less centered around the green or blue curve, respectively?

Dear Reinhold,

Thank you for revising the manuscript again. In light of the second reviewer’s comment on language, I decided to read your manuscript thoroughly once again. I have still found a few sentences that are difficult to understand and some other very minor issues. Please find attached a commented PDF file.

Thank you very much for your collaboration and patience in this somewhat lengthy review process!

Best regards,
Stephan

Dear Reinhold,

I think your manuscript is now ready for publication. I've just spotted a very minor issue. Line 31: Please replace 'ecological' by 'economical'. This modification somehow got lost.

Best regards,
Stephan