Thanks for your constructive and comprehensive response to the review comments. Please proceed with the revisions you've outlined.

thanks for your work revising the manuscript. As you'll see, there are a few more minor comments that the reviewers feel would improve the work. Reviewer #2 notes that your original response did not address their remarks in a point-by-point format as requested, and there are a few places where it's not clear how you have responded. I'd appreciate it if you could remedy this, and produce for the record a clear list of replies to each of the reviewers comments from both their original reviews and these latest ones. Thanks.

Dear Xavier et al,

Thanks for your further consideration of the referees comments, and the clarification and repeat of the point-by-point replies to their comments. Most of what you've done is fine, but I'd like to continue working on one point a little. Reviewer 1's recommendation for specific issues to be addressed in the Discussion sounds very useful to me, and whilst what you've put in the paper in response is perfectly reasonable, I think it still takes some reading between the lines for a reader to extract from that the take-home information that the Reviewer thought should be available.

The reviewer is asking for some discussion - or at least informed speculation - of how far the performance of various model types in the climate of the recent past can be translated into an expectation of their performance under future, much warmer conditions, with a specific consideration of the case of computationally cheap PDD models which may be over-tuned to present conditions. I'm inferring from what you've put in this revision that you think that model performance during the 2012 summer is a sufficient test of how they will perform in the future, and that runoff bias during the GRSMBMIP evaluation period is a better indicator of how accurate an SMB model will be in future than anything in the formulation of model physics. However, a reader is left to extrapolate for themselves to this sort of conclusion from what you've said, and I may have got your meaning wrong.

Could you add in a sentence to be really much more explicit about your conclusions from GrSMBMIP in the context of the specific points about model's likely performance under future, warmer climates that are being queried by Reviewer 1?

thanks,

Robin

Thanks, I think that's sufficient clarification of the point in question. I think this should go forward to publication now.