After carful evaluation/consideration of the two reviews, the authors revisions and response and the revised manuscript I took the decision that this paper should be accepted and published in TC as it is.

This was not an easy decision due to somewhat different opinions of the reviewers.
Review 1 was very positive and explains well why this paper is a valuable contribution to TC (in terms of novelty, significance and quality) and I do in general agree with this reviewer. However, some suggestions for improvement/clarification were also made by reviewer 2 which were fully addressed in the revised version and commented in detail in the author response.
Review 2, was however rather critical and rejected the publication or recommended a major rewrite with a rather different focus of the paper (therefore ‘new’ paper). Although the relatively simplistic modelling approach was in general accepted (as it captures essential physics), reviewer 2 was of the opinion that for this case of relatively ‘minimal’ data constraints for the modelling one should be more careful and not promote a quantitative ‘predictions’ of future behaviour of the chosen surge glacier.
I very carefully considered the comments of reviewer 2 and in principle agree with the concern of reviewer 2. However, as an editor I did not think there were sufficient objective grounds to reject the paper as:
i) the paper was very clear and transparent about the assumptions and simplifications made
ii) the uncertainties and simplifications made were very well discussed and also analysed quantitatively and in detail with sensitivity experiments (as also mentioned by reviewer 1)
iii) and the data constraints were actually not that ‘minimal’ and the modelling in my opinion not that poorly constraint compared to many other modelling studies.
iv) And the investigation of effect of (prescribed) surges in context of a general warming and glacier retreat in itself new and very interesting.

Thus overall I thought the presented modelling study is certainly a valuable contribution to TC, the approach clear and robust within the given/described limits. I also took notice and generally agree with the first author response defending their paper/modelling approach.
However, I insisted as editor that the main points of reviewer 2 should to some degree be taken into account in the final version such as i) a clarification of the data constraints used and effect of data uncertainties, ii) to flag the uncertainties of the ‘predictions’ made and make clear that they should not be read as an ‘exact prediction’, but instead as indicative of a possible future scenario and iii) adding that further testing of the modelling approach with more/better data constraints is needed.
The revised version of the paper fully includes and addressed these points and think the made interpretation and conclusions of the paper are fully supported by the undertaken and modelling study.

Therefore, after carefully considering all the available assessments, responses and checking the revised document, I am convinced that this paper is a valuable and solid contribution to TC and recommend to accept it as it is.