Dear authors,

thanks for the online replies. Both reviewers suggest minor revisions, which however then do ask for several specific clarifications and also for better arguments in some cases.
To me a very important one you need to address is the usefulness of SPI-1 forecast, when almost every analysis and monitoring of drought impacts is currently done based on longer aggregation times (SPI-3-6 at least). Please address this issue very carefully.
Also, I think the question by ref 1, whether forecasting improves after drought onset, I think would make the paper much stronger if there was anything you could draw on this question from your analysis. It also picks up on the question I raised earlier and would have liked to see addressed, i.e. whether a numerical weather forecast derived SPI-1 forecast is better than the probabilistic ones that have been published in other studies. If this could be shown, it would add a strong point and impact.

I am looking forward to reviewing the revised paper.
Regards,
Kerstin

Dear authors,

thanks for the revised manuscript. I have to request a few minor improvements before publication:

Page 3 -> merge the mini-paragraph on definitions with the next or possibly the first three paragraphs. The link of these definitions to the work are not clear to me. Why do other definitions than meteo drought matter for this paper? -> Tie in better

Both reviewers and myself asked to discuss the issue of only looking at SPI- 1 in the Discussion section. The condensed bit of text added to the introduction is not sufficient to address this major concern. It also does not fully reflect your more detailed response letter.
-> Revise this aspect also in the discussion section! There is still also not much embedding into other work. If you argue with the usefulness for operational application, you should review and point to, which operational systems in the world already use a one month forecast and how and how your study may influence progress in this respect. Quickly browsing, I found E.g. http://droughtmonitor.unl.edu/SupplementalInfo/Forecasts.aspx.
And I am sure a few of these global applications also do this, no? So how exactly will your study help them improve?

New equations for skill scores:
->Add references to the equation numbers in the text

Page 19 new paragraph:
a) first sentence:-> remove “global” or find another descriptor, but the application was neither ‘global’ in geographic sense, nor in the sense of an all-encompassing assessment (but rather narrowly limited to SPI-1).
b) should be ‘other methods’
c) Since it is rather unclear what is meant by ‘other methods’ and other ‘atmospheric predictors’ (which, whatfor – should it be drought indicators?), I suggest to swap the order of the two sentences somehow to make this relation to other approaches clearer.
In fact, since the last part then goes back to SPI-1 (after mentioning other predictors (of drought?), it may be more useful to have the new part last. Please revise this last paragraph for better logic and more clarity.

Regards
Kerstin