|Review of revised Quiquet et al. manuscript: Online dynamical downscaling of temperature and precipitation within the iLOVECLIM model (version 1.1)|
Thanks to the authors for revising based on my comments and also comments of Reviewer 1. I’d suggest a few more general revisions:
-Abstract: I don’t think it’s correct in this case to state the scheme conserves energy and moisture - particularly the former. Conservation of energy entails much more than conservative interpolation of temperature. For example, can iLOVECLIM, that includes the downscaling scheme, demonstrate a globally closed energy budget, including, e.g., latent heat transformations (such that the net energy change in the system is equal to the net TOA flux)?
-Abstract (and elsewhere): In the Reply to Reviewers the authors write (note paraphrasing) “Regarding the SMB, [grid imprinting issues] presented in the manuscript are not suitable for an ice sheet model forcing” and “Due to the imprint of the coarse resolution model into the current downscaled fields, the latter cannot be used directly into the SMB and need additional steps beyond the scope of this current study…”
Yet in the public-facing manuscript, the authors repeatedly state: “Foreseen applications of this new model feature includes ice sheet model coupling…”. There is an apparent disconnect here, between what the authors honestly think their model is capable of, and how its capabilities are described in the manuscript. The authors need to be more explicit with readers about the applicability of their developments, *as they stand* - for example, if the work presented here is considered an interim benchmark on the way to something that is scientifically useful for, e.g., ice sheet SMB science, this needs to be clearly stated.
-On grid imprinting: I think the reasons for actual grid imprinting are still not described clearly enough for the general reader. Regarding temperature, for example: “This is because our downscaling mostly redistribute (sic) the temperature of a coarse grid point according to the sub-grid elevation starting from the coarse grid information” - is an ambiguous sentence that leaves the reader puzzled (and does not imply that this problem also extends to precipitation, as noted in the Reply to Reviews). Regarding precipitation: even if you took winds into account, it would still be true that the main (only?) effect of downscaling is to redistribute precipitation within a native T21 grid cell. Thus, T21 grid cell imprinting will almost certainly still occur across the location of T21 grid cell boundaries in downscaled precipitation.
As above, I feel the authors should be more transparent with describing the caveats of their approach, particularly with respect to remaining grid imprinting - just so that potential users aren’t surprised/frustrated when actually applying the scheme (or run into issues when they submit science using the scheme to peer review).
-There remain a fairly large number of grammatical errors that the authors could sweep for prior to any official GMD proof-read.