|Thanks to the authors for the detailed and extensive response to the reviewer comments. The response addressed as promised all points and clarified several concerns. Nevertheless, several concerns remain, and others occurred during re-reading the manuscript. |
Before I comment on the comments to reviewer#2 , I would like to pick up a comment to reviewer#1. Regarding the novelty and the comments made to Reviewer#1: I see your point about the model differences, but I am not sure if this point is clearly pointed out in the presentation of your study. If I understand the authors correctly there are two major differences between the here presented study and former studies:
1.) The here used model approach is a process-based model approach, while ethe former studies are mainly data driven approaches (including more details on e.g. the management).
2.) The finer spatial resolution.
Please make this clearer in the manuscript. I understand that the comment seems to be unpleasant, but your response is only mentioned in comments and not reflected by changes in the manuscript. Considering the list you mention, there are models like DAYCENT and Landscape-DNDC includes, which are also state-of-the-art models on croplands. The combination of grassland data and cropland data might be new, but is not the new innovative step that would be used to address this comment. Please characterise the novelty of the study more clearly in the manuscript, as reviewer#1 seem to miss this (and my comments went in a similar direction).
Concerning the location specific differences in the results, I would like to thank the authors for their explanations. I can follow your explanations. I do not think that the representation of fertilizer application in Germany is correct (too high in the North-East (considering a fertilizer shortage during the GDR times in the 80s and the low fertility of the soils) and too low in the North-West (this might be due to the difficulties to estimate the amounts of applied manure)). The fact that the sharp changes at the political boarders are visible shows the statistical flaws of this data set. A critical view on the SOC values let also assume a fairly high SOC stock for croplands. These are comments to the data sets and not on the study here. The comment on the study concerns a more detailed elaboration of these facts in the discussion. Serval maps allow to recognise the political boarders of countries, which is an indication for the impact of the statistical aggregation of the used data sets. A more detailed characterisation about precipitation driven emissions (South) and more temperature driven effects (North) would be great. For me the input data seem to be the main drivers for the simulation result. To underline this, I would like to pick up your argument with the organic content in the soils. First, do the authors think that these high SOC values are realistic for cropland soils? There are no cropland specific SOC values available, so it would be good to add a short check of references that back-up the assumptions for the high SOC values in Eastern Europe (this is especially doubtful in Poland). Are organic soils excluded from the data set? The authors might argue that looking in the difference or changes due to differences in weather data are the main focus, but some of these “doubtful” areas show the strongest response.
I disagree with the conclusion. There are a couple of points:
- Lines 871-873: this is exactly what was discussed and the authors agreed on. These are not absolute values and absolute conclusions are hardly possible. The simulation results show that the changed climate show a minor impact on the simulation results.
- Lines 875-877: I think this description is mis-leading. These are managed systems, which need to be caracterised by the NBP rather than the NEP. Both systems are there for producing yield and hay. The argumentation yield and hay removal changes the system to a sink sounds a bit odd. Even though the description technical not wrong, it would be great to have a different formulation here.
- Line 878-879: this is not a conclusion, but a discussion point. Did you test the changes for leaving more residues on the field?
- The overall conclusion is for me that the input data and their uncertainty are a more relevant driver for changes than the climate. Only at the end of the century the climate affects strong changes, which might show the potential to show a larger impact than the management uncertainty.
In figure 5, S5 and some others it is not clear enough that the difference between the maps is only the irrigation. I also do not think that all these maps are necessary. The difference between the irrigated and non-irrigated maps is too small to recognise. I think it is an important and relevant finding, but one map with irrigation as example and the rest mentioned as showing a similar los change would be enough. In figure 12 the text is not readable (N2O emissions).
The authors mention changes in the growing season of spring crops. Considering that the growing season was derive by phenological models, I would assume a reduction of the growing season as well, as the vernalization would be extended due to increased winter temperatures. Was the vernalization included in the models? I understand the author comments on the second growing season. However, considering the shorter growing seasons for some areas, make a second growing season a realistic option and will affect the overall balance. This should be discussed.
Lines 520-533 and figures 10 and 11: I am not sure, if this is a good way to discuss this aspect. The yield needs to be removed to provide a more correct estimate for the “real” input into the system. It is possible to analyse the NEP for the changes, and, assuming the harvest index will be constant, the relative changes will stay the same. However, I think the NBP would be a greater interest, as it indicates if the cropland/grassland systems are an overall source or sink. As I assume that the authors might have other thoughts on this, the yield and/or the NBP should be added somewhere (also the supplement would be fine with me).
Lines 784-785: Please sort out the abbreviations and its definitions. First, NGHGE is not the balance, but the exchange. Second, the NGB is the GHG balance, not only C forms (I prefer fluxes). Third, the removed residues needs to be added, while the remaining residues stay in the system.
It looks like the maps are created by a GIS tool. Please change the legand numbers to suitable step-sizes and remove the high number of positions behind the comma (e.g. in figure S1).
The nature of a review is often the critical sound and comments. I also would like to highlight the positives, as the responses to the comments were very good and especially the specification about the spin-up and the main objective. In terms of critical analysis and novelty I would like to have seen more changes in the discussion and conclusion section (more parts of the comments moved to the manuscript). I thank for the discussion and looking forward to the comments.