|The authors have provided a detailed and thoughtful response to my initial review. Looking back, my primary concerns were that 1) the numerical experiments were not reproducible because the source code was not well documented, and the coupling not described in detail, 2) that the asynchronous runs might mask numerical instabilities, 3) the design criteria are unclear, and 4) that the summary results relied on metrics and plots that were not that informative. |
Issue 1 has been addressed by making the source code available online, and by providing additional text in a number of areas. In particular, the stating inputs and outputs of BrAHMs are instructive. I no longer believe this is an issue.
Issue 2 was closely tied to 3 - in the case of a ice sheet systems model, I had questions about the interaction between ice sheet and hydrology components. The authors have taken care to scope the paper towards the development of the hyrdrology only, with some tests of its function in asynchronous experiments. Rereading, I find this reasonable, and am satisfied with the outcome. Additionally, the authors made efforts to demostrate the stability of the model in a number of ways, for example, figures 2 and 4 are excellent.
Issue 3 - by adding four paragraphs to differentiate between BrAHMs and other large scale ice sheet models including hydrology helps to clarify the importance of this work, and makes many of the questions raised in the original review less relevant.
Issue 4 - Both reviewers made comments on the figures, and I'm glad to see that some changes were made. I also liked the paragraph that acknowledges that local sensitivity may be obscured by looking at bulk results. I'd like the authors to consider, a final time, the possibility that the information density is low in figures 1 and 3 - and that the visual representation obscures potentially interesting details in the spatial distribution of water thickness and pressure. Both reviewers note this. But, if the authors feel strongly, I'm not going to get in the way. Finally, please read a little about colorblindness and the 'jet' color map. I get the sense that these figures are mostly OK because they use only 10 unique colors (good!), but you'd hate to lose audience because of something as superficial as the colormap.