Review of “Undercatch corrected gridded precipitation data to improve hydrological modeling in high-alpine orography” by Philipp Maier et al.  

Maier et al. used weather station data in Austria to create correction factors to adjust for the well-documented wind-induced undercatch that precipitation gauges are suffering from. They further developed a method to expand those correction factors and present a gridded data set, adjusted for undercatch. The results were validated with independent measurements from two catchments in Austria and the authors could show that a general more realistic representation of the precipitation were achieved.  

While the idea of developing a gridded data product with precipitation data adjusted for undercatch is not new, the development of a consistent and complex method for it and its application to a new region in this paper, justifies the publication of this paper in HESS. The study appears well prepared and the detailed verification of the results at the end, and the discussion of the limitations witness of a thorough process.  

I do think, however, that this thoroughness is not reflected in the presentation quality of the paper. There is a need for major improvements in the description and explanations of the underlying data and calculations which will also increase the scientific quality of this study.  Please see my major comments on this. In my opinion, also the use of references is insufficient repeatedly throughout the study. Some references seem cited slightly out of content, and some do not clearly difference what part is the authors’ contribution and what is taken from the citation. I will give some examples at the end but encourage the authors to carefully re-check all their references and possibly adding furhter studies which work on similar topics in different regions before resubmitting a revised version.

Major Comments: 

Section 2.2 – Weather Station Data 

Line 125ff. This entire section is missing important information.  

• It is unclear to me what exactly you mean with “semi-automated” weather stations. What of them is not fully automated? Do all stations have all the parameters you mention? What about their location and elevation? Much later, in your result section line 277 you state that wind and precipitation sensors are not always co-located. Also, the elevation of the precipitation data will become an important criteria in your results – this information is clearly missing here. 

• You should describe and explain what gauge-types are in use already in this section. You later mention that it is mostly tipping buckets (line 159, section 2.4.), as a reasoning for your selected CE transfer function (which then was from the paper not describing tipping buckets, see comment for section 2.4), but also name in line 211 that you want to account for the use of gauge totalizators (possibly as base for the gridded data set) without any further explanations. Please gather all the information about your station data here in this section. 

• Another interesting (and missing) factor is which time periods of data are you analyzing? You mention that you require at least 10 years of data, resulting in your selection of 261 stations – but do you analyse 10 years – do all stations cover the same 10 years? 

Section 2.4. 

• Lines 155-160, including equation (1) : You present the CE transfer function from Kochendorfer et al. 2017a (which is the correct reference for the presented equation at this point; please also see the reference comment for line 89-91). However, this equation was not, as stated in the next sentence, developed for tipping bucket gauges, but for weighing gauges. The equation to be used for tipping bucket gauges would be from Kochendorfer 2020 (also in your list of references). It is, however, slightly different and requires different parameters. Depending on the gauge type in your dataset, which needs to be mentioned and described in section 2.2. (see earlier comment), you may want to change the equation, which would demand a re-calculation of the gridded data set as well. If you are deciding to keep this equation, please double check your constants. You write α=0.6023, while I find α=0.06023 in Kochendorfer et al. 2017a, Table 2.  

• Line 165.  What is your rationale behind aggregating your exposure factors to monthly averages? Both temperature and wind can vary a lot within a month, thus creating possibly the whole range of corrections factors. Off course local climate will be seen on the average – but this rather extreme aggregation step (30min to 1 month) needs more explanation, as you lose a lot of information you just have gathered. You could show the distribution within the months (are they normally distributed, what is the variance, how often do you have extreme values which may be overrepresented in your average…).  

Section 3.2 Description of Figure 6, Lines 256 ff. 

• Please be more concrete here and describe your results in more detail. The term “dominant relationship” means in your case a whole order of magnitude, factor of 10 at least for the summer months. It is not so clear, however, if this factor is dominant for the winter months, as all winter curves seem to gather under a multiplier value of 5, which would be in the order of magnitude with the two other features.  Also, for me it is not obvious why the elevation dependend factor should be much larger during summer than during winter – I was expecting it to be the other way. 

Minor Comments 

Section 2.2 – Weather Station Data 

• Line 125-126: Readers outside Austria may not know what Geosphere Austria is – please explain. It is listed here in the format of a reference, but not listed in the reference list, that can be solved by spending a few more words on it. 

Section  2.5 

• Lines 190-194: Please explain your definition of “exposure”. I do think that terrain elevation may also be a proxy for wind speed.  

Section 2.6 

• Line 212 – here you mention you want to account for the use of gauge totalizators in the gridded precipitation data set by Hiebl and Frei, 2018. Is there a difference in the gauges used for the gridded dataset and those for calculating the correction factors? In which way?  

Section 2.7 

• Line 230. Include “Figure” before 4 

Section 3.2 

• Description of figure 6: For the 5km-exposure (6b) one month is clearly different but not mentioned.  Please also reconsider your choice of color. For me, it is difficult to decide which of the “blue months” is the one showing the different behavior. Further, the 51-km exposure in figure 6c shows a significant different behavior for winter months which should be mentioned and discussed in the text.  

Section 3.3 

• Line 341-343 – The changed seasonality in your two catchments is a very interesting result and should be highlighted more.  

• Line 351 – I guess here you mean that the pure dependency of the transfer function on wind speed is not enough in complex terrain, as first the use of your ETP gives you the wished results? I do in principle agree that complex terrain is very challenging to adjust, but I wonder if it is simply because it is more difficult to find a wind speed measurement representative for the conditions at the gauge’s orifice as a nearby wind sensor may capture different conditions and thus give a wrong correction. In any case, please clarify your argument.  

References. 

Please note that I didn’t check every single reference. After checking a few citations, I noticed repeatedly slight mismatches between the citation in your text and the content in the paper. Here are some examples. But I advise having a thorough check of all your references. 

• Line 26: Formayer et al., 2023 presents and publishes the dataset SECURES-Met, developed for electricity model applications. Its hydropower potential parameter is based on river discharge measurements. It mentions also challenges connected with topography, but I don’t think it is the right source for supporting the statement that reliable precipitation estimates become increasingly critical and have impacts on the hydropower generation in alpine regions. The word precipitation is mentioned only once in the entire article.  

• Line 47:  I think it is appropriate to mention other gridded data products with adjustments for precipitation in complex terrain as for example Norway and Switzerland. I added two to sources to consider, but there are possibly others:  Lussana et al., 2019: seNorge_2018, daily precipitation, and temperature datasets over Norway, Earth Syst. Sci. Data, 11, 1531–1551, https://doi.org/10.5194/essd-11-1531-2019, 2019;  Isotta et al., 2019: Long-Term Consistent Monthly Temperature and Precipitation Grid Data Sets for Switzerland Over the Past 150 Years DOI:10.1029/2018JD029910; 

• Lines 89-91: inconsistency/typo? Here you are referring to Kochendorfer et al. (2017b) as source for your undercatch adjustment method, while in your methods chapter, lines 155-158, you state that you are using the undercatch correction developed from Kochendorfer et al. (2017a). Please be consistent and also check my comment on the use of the equation from Kochendorfer et al. (2017a) under major comments.  

• Line 95: It reads as you are introducing the "exposure-dependent penalty" by citing papers Kochendorfer et al (2017a), Hiebel and Frei (2018) and Gnann et al. (2025). Kochendorfer et al. (2017a) mentions only that complex terrain can introduce a higher uncertainty when using adjustment functions, Hiebel and Frei (2018) are mentioning only once that undercatch may be part of their data and Gnann et al (2025) mentions a lot of topography's influence on the development and distribution of precipitation, but also here undercatch is only mentioned once. Neither of these papers is suggesting any exposure-dependent penalty as this sentence and citation suggests.  It rather seems, that the ETP, is your invention – so I guess, here a slight reformulation of the citations will do.  

• Line 120: Please reference where the numbers of the glacier coverage for your catchments are taken of. Also, I was wondering if you have newer numbers than from 2003, as you at other places cite Hugonett et al. (2021) which gives you a timeseries of glacier volume (shown in figure 14)   

Section 2.3. - Historical Climate Data,  

• In this entire section you refer to numerous other data sets which you are using and explain how some of the data you are using were derived or processed. Please clarify your writing to clearly indicate which of these processing steps were done by you within this study and which by the creators of the cited datasets. A representative example: Line 141, citation of Jiawai Zhuang et al., 2025, which referes to a dataset (zenodo-DOI) but is cited after the sentence: “After applying the undercatch correction, both the raw and corrected precipitation data were patch-interpolated to the study regions” - Who did the patch-interpolation and what exactly is in this dataset? Its name, when opening the link: “pangeo-data/xESMF: v0.9.2 - Third time's a charm” does not give enough information to the interested reader. Please revise the citations in this entire section to clarify.  

Section 2.7

• Figure 4: There is no citation for this figure, which I assume is taken from one of the COSERO-documents? Please add. 

• Line 250: Paper Wolff et al. 2015, alongside with two others , is cited for the statement that precipitation lapse rates may be plateauing at ridges above 3000 m - and while this paper is presenting results from a 1000m-high alpine station, it doesn't include any discussion about precipitation undercatch in higher altitudes or their "lapse rate". I did not check the two others here – please do so yourself. 

Section 3.2.

• Lines 312 and 318 – you are referring to the same three studies twice for almost the same thing, the dependency of undercatch on the terrain elevation as the major descriptor (and its seasonality added in the second time) without any concrete example/comparison or differentiation. At the very least, strike one of the sentences. I do, however, suggest a more thorough discussion here, especially as all these three studies seem to discuss possibly adjustment methods for precipitation in the alpine region and you clearly take inspiration from those. How is your method differing and what is similar? Are any of the methods superior and why?