We are thankful to the reviewer for taking the time to review this manuscript. We have answered all the comments of the reviewer below.  

The reviewer comments appear in bold, our answers in normal font, and changes to the manuscript in italics. 

I suggest then to the authors to resubmit their paper in GMD (Geoscientific Model Development) which fits a lot of better with this kind of papers.  

We disagree, this is not purely a model description paper. A large paper describing the CryoGrid model is already available in GMD (Westermann et al, 2023), and describes the model in much deeper detail than we do here. A journal like Earth System Science Data could have been fitting, but they do not allow model output descriptions from climate model/land surface models. At the suggestion of another reviewer, we have now restructured the introduction to make the research aims clearer. 

Before resubmission, I would just like to highlight that I don’t see the interest of the addition of AROME-ARCTIC as forcing of the snow model in this paper. With respect to CARRA, AROME-ARCTIC is an operational product forced by IFS and can not be used in any scientific relevant papers knowing that ERA5 is available in real time with a delay of 1 week with the present date and only using AROME-ARCTIC forced by ERA5 will be relevant here for me.  If the aim of the authors is to show the interest of AROME-ARCTIC used in forecast mode, I suggest in this case to change the focus of the paper to this aim by considering only 2016-2021. But, discussing the recent changes over 1991-2021 + evaluating the interest of AROME-ARCTIC over 2016-2021 in the same time decreases the interest of the paper for me. 

We disagree that AROME-ARCTIC is not useful because ERA5 exists. ERA5 is a useful product, but the resolution is too low for simulations of glaciers on Svalbard (30 km vs the 2.5 km of AROME-ARCTIC). AROME-ARCTIC uses the same model as CARRA with a very similar setup and assimilates most of the same data. It therefore makes sense to evaluate the two products in the same paper, particularly since AROME-ARCTIC simulations have already been used in previous studies. However, we try to more clearly separate the results based on CARRA from those based on AROME-Arctic, such that readers who are not interested in AROME-ARCTIC, can simply skip section 5.4 of the results. 

Some additional minor remarks: 

The use of the words "climatic mass balance" or "glacier mass balance" is ambiguous here for me as both mean surface mass balance. I suggest than to use SMB everywhere in the paper. 

There is a difference between the total mass balance, climatic mass balance, and surface mass balance, and we therefore would like to keep the distinction. The surface mass balance (SMB) quantifies the mass fluxes between the atmosphere and the glacier at the surface, and within the current year’s snow layer (refreezing within the annual layer). This is what is measured by in-situ glaciological observations. The climatic mass balance (CMB) additionally accounts for mass changes below the last summer surface and can therefore be simulated with a model like the one used in this study. Total glacier mass balance is the sum of CMB, basal mass balance and frontal ablation (i.e., subaqueous melting and calving). The latter term only applies to glaciers terminating in the ocean or freshwater lakes. The terminology used in our manuscript follows that recommended by Cogley et al (2011) 

We have made this distinction clearer in the introduction:  

L80-82: CryoGrid simulates both the surface mass balance (SMB) and the climatic mass balance (CMB). The surface mass balance quantifies the mass fluxes between the atmosphere and the glacier at the surface, as well as refreezing within the annual layer. The SMB is what is measured by in-situ glaciological observations. The climatic mass balance additionally accounts for mass changes below the last summer surface. For tidewater glaciers, CryoGrid, however, cannot calculate the total glacier mass balance, as this is the sum of CMB, basal mass balance and frontal ablation, and the latter cannot be determined from an energy-balance model. 

In Section 5.6, runoff is given in GT/yr while SMB/precip is discussed in mWE/yr. I suggest to use the same units (m WE/yr or GT/yr)  through all the paper. 

This is a good point. We have given the runoff in Gt/yr, as this unit is often used for runoff, while m w.e./yr is often used for the SMB. However, in order to better compare, we have now added the runoff given as m w.e./yr in both the runoff figure and in the text, so it is easier to compare. We still, in some cases, give the runoff in Gt/yr in addition so it easy to compare with the runoff from other studies. 

What is the interest of Section 5.4? It is mainly a comparison between CARRA and  AROME-ARCTIC forcing but what are impacts of the seasonal snowpack recent changes on SMB or climate? A comparison with observations could be useful if the aim is to validate it. 

The idea was to show how the seasonal snowpack has evolved over the last 30 years, but you are right that it is maybe not the most relevant for this paper. We have removed the section from the revised manuscript. 

In conclusion, increases in land runoff is mentioned without having discussed more in depth changes in precipitation (rainfall/snowfall). Discussing land vs glacier trend will be useful here in addition to use the same units for precipitation and runoff. More in general, what is the interest of discussing here the land runoff changes? What are the impacts on ocean? Are there some observations confirming this modelled estimate? 

We have now added a bit more discussion on the changes in precipitation (rainfall vs snow): 

L290-298: The average precipitation over Svalbard is 0.62 m w.e. yr−1 . There is a small but significant trend in the average yearly precipitation of 0.05 m w.e. decade−1 (p < 0.01). There is a larger trend in the precipitation over glacier-covered points (0.06m w.e. decade−1) than non-glacier-covered points (0.03m w.e. decade−1). Although there is no significant trend for all areas of Svalbard, there is a positive trend over e.g. Austfonna, Vestfonna, and North Spitsbergen. The largest trend is over NE Austfonna of 0.17 m w.e. decade−1. Over the investigated period, on average 90% of the precipitation fell as snowfall. There is a significant decreasing trend in the ratio between snow and rain, with the percentage of precipitation falling as snow decreasing by -2.0% decade−1 (p=0.01). For glacier-covered points, 95% of the precipitation falls as snow, with a significant decreasing trend of -1.3% decade−1 (p=0.02). Over non glacier-covered points, however, 85% of the precipitation falls as snow, with a significant decreasing trend of -2.8% decade−1 (p=0.01). 

The interest in discussing land runoff is to compare it to the runoff from glaciers to get an idea about the runoff contributions from seasonal snow. Runoff from seasonal snow generally occurs earlier in the year, so it therefore affects the runoff into the ocean. In addition, the effect of the ocean is different depending on if the runoff comes from a tidewater glacier or from seasonal snow or land terminating glacier (e.g. in terms of nutrients, light pollution, fjord circulation). This is, however, not the focus of this study. But we have now added a few lines in the conclusion to highlight the significance of runoff for the ocean:  

L558-59: The timing and amount of freshwater runoff from Svalbard has important implications for the ecosystems in the surrounding fjords. Changes is freshwater discharge affect a wide range of physical, chemical, and biological processes, including e.g. fjord circulation (Carroll 

et al. (2017)), light availability (Hop et al. (2002); ARIMITSU et al. (2012)), water biogeochemisty (Wadham et al. (2013); Bhatia et al. (2013)), and marine primary production (Juul-Pedersen et al. (2015); Hopwood et al. (2020)). Freshwater from tidewater glaciers may affect these processes in a different manner from seasonal snow runoff or runoff from land-terminating glaciers, 

and it is therefore important to quantify the amount of different types of runoff. 

We had not included any evaluation of the runoff, as very few observations are available from glaciated catchments. Now we have included a comparison with two partially glaciated catchments (Bayelva and De Geerdalen) as section 5.2.3.  

Citations:  

Cogley, J.G., R. Hock, L.A. Rasmussen, A.A. Arendt, A. Bauder, R.J. Braithwaite, P. Jansson, G. Kaser, M. Möller, L. Nicholson and M. Zemp, 2011, Glossary of Glacier Mass Balance and Related Terms, IHP-VII Technical Documents in Hydrology No. 86, IACS Contribution No. 2, UNESCO-IHP, Paris. 

We are very grateful for the constructive comments and suggestions provided by the reviewer that have significantly improved our manuscript.  We have included all suggestions for changes and have outlined response the comments in the attached file.

We are very grateful for the constructive comments and suggestions provided by the reviewer that have significantly improved our manuscript. We have added point by point responses to all comments in the attached file.