Reviewer comments for egusphere-2025-636

This study performed a multivariable calibration on a hydrologic model, that went beyond the traditionally calibrated discharge to include other components such as ET and groundwater flow. I think the authors did a good job incorporating my previous comments about bringing in more existing literature and explaining the “why” of the study. I really like the improvements the authors made. With these improvements, I believe there is now work to do to highlight and clarify the advancements of this study beyond the previous hydrology and multivariable calibration studies mentioned. I also feel that the results should go beyond the performance of the calibrated model to testing a hypothesis or comparing with other calibration approaches, to help support the statements made in the conclusion. I provide suggestions to augment the paper in this regard.

Major comments:

Page 4, line 80. I think you need to elaborate more on what makes the calibration a new approach. You mention earlier in the introduction other modeling studies that have incorporated ET, groundwater, and soils. Is the novelty here just that you used these in a new combination, or is there something more specific you can describe, such as with the use of PTF’s? You also mention the “novel approach” on line 94. I think clarifying the specific novelty in more detail would help show how this goes beyond previous work.

Page 4, lines 91-95. I suggest adding a scientific research question and hypothesis, again to help make it clearer what the advancement of the study is, and make the novelties more distinguishable. This would help clarify what is tested in the study and what the focus is beyond the methods. For instance, what is “more accurate” in comparison to? Is there any hypothesis you could develop and test with the PTF’s in the calibration?

Page 24, line 452. At the end of the results, I can’t help feeling that this section is cut off too soon. I’m not convinced that stopping at the model performance evaluations is sufficient. This could go back to my comment about having a hypothesis to test, to go beyond just describing a method. As an idea, would it be possible to add a comparison with a model calibrated with a traditional, discharge-only approach, or with other multivariate approaches from the literature? In terms of performance, parameterizations, land use and climate change simulation, etc. This could help provide quantitative evidence showing why a modeler should use your approach rather than other common approaches, and what the improvement could be for their work if they adopt it. Or, as another idea, clarifying and testing a hypothesis with PTF’s could help accomplish this. I think it would be beneficial to update the abstract to include a sentence of the findings beyond performance evaluations (lines 16-19) as well.

Page 28, lines 531-534. I like this statement about the importance of giving attention to PTF’s and how this is the first study to use them in calibration. I think this should be elaborated on earlier, such as at the end of the introduction, and could be the basis for a novel hypothesis about this narrower focus.

Page 30, line 580 to page 31, 594. I think the conclusions section still has a way to go. The first paragraph is essentially summarizing the previous literature of the introduction. I suggest rewriting it to be more about the learnings from this study beyond that. The second paragraph discusses reasons the new approach is better than previous approaches. However, this is largely conceptual, as the new approach has not been compared with previous approaches in the results. I think adding a quantitative comparison with a different calibration approach or something similar to the results as I suggest could help provide support for these statements.

Title. If I’m understanding right, the current title describes doing something that other studies have done in the literature review of the introduction? (multi-variable process-based calibration of a behavioural hydrological model). If so, I suggest making the title more specific to what this study does that goes beyond the previous literature, similar to my comments above. For instance, it could be an answer to a research question or hypothesis that is tested, such as with PTF’s.

Minor comments:

Page 6, line 124. How does the model handle snowmelt? It may be worth a brief explanation if snowmelt is relevant to the study period and parameters, since it’s later mentioned in lines 280 and 570.

Table 1. Would it be possible to add another column to this table that has a brief description of the PTF’s for each row? For instance, something to help explain why the dominant runoff process patterns (interflow, surface runoff, deep percolation, etc.) could be different among the PTF’s in section 3.3.

Page 18, lines 350-360. As these performance metrics are commonly used, I suggest to remove these equations and descriptions, and just direct readers who are interested to the references.

Page 29, line 540 to page 30, line 579. I got a lot out of this section. I thought it answered well the questions I had about replicability and transferability from my previous comments.

I wish the authors the best with this work and their future endeavors.

In this manuscript, Heuer et al. present a multi-step calibration approach and test 12 parametrizations of pedotransfer function for plausibility with regard to spatial patterns of simulated process dominance.

I like the approach of the authors and think that it is a valuable contribution to process-based model calibration. I have only a few comments.

Comments:

It is not always clear whether the authors are analysing evapotranspiration (ETa) or evaporation. It is important to clarify this and to make clear that the comparison between MODIS data and simulation results is valid. See for example L. 10 “Evapotranspiration parameters based on MODIS evaporation data”. See also L. 178-180.

The selection of the hydrograph efficiency metrics is not convincing. NSE and R2 are conceptually very similar. There is no advantage in using R2 in addition to NSE. Table 5 shows that a lower NSE results in a lower R2 and vice versa. So, R2 is not needed. I suggest adding a complementary metric e.g. PBIAS to capture another part of the hydrological cycle. Another option could be to use the three components of the KGE separately (as KGE-beta represents the same hydrological behaviour as PBIAS).

L.371-374: This part needs to be reformulated. Three sentences mention that PTF 9 and 10 are not valid. I suggest to make one clear statement.

Table 5: Use consistently two digit numbers. For PTF1 and KGEval three digit numbers are shown.