the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 03 Sep 2026
ALPAKAS: Alpine hydrographs and catchment attributes for karst springs
Abstract. Approximately 40 % of the European Alps consist of karstifiable rock. Due to their distinct recharge, storage, and discharge properties, karst aquifers are key components of the hydrological cycle and water supply in the Alpine region. However, large-sample datasets enabling comparative and data-driven analyses of karst spring discharge are currently lacking. Here, we present ALPAKAS (ALPine hydrographs and catchment Attributes for KArst Springs), the first large-sample benchmark dataset specifically dedicated to karst springs. ALPAKAS comprises 242 daily and 183 hourly quality-controlled and consistently preprocessed discharge time series from karst spring stations distributed across the EU Strategy for the Alpine Region (EUSALP) area, with stations located in seven countries. The dataset is complemented by both existing catchment delineations and circular catchment approximations, catchment-aggregated meteorological time series at multiple spatial scales, and a comprehensive set of static catchment attributes describing topography, positioning, meteorology, land cover, soil properties, and hydrogeology. In addition, the dataset includes karst-specific indices directly derived from the karst spring hydrographs.
ALPAKAS is available via Zenodo at https://doi.org/10.5281/zenodo.19329694 (Joger et al., 2026) under an open-access licence and is accompanied by comprehensive documentation. By providing a harmonized, large-sample compilation of karst spring catchments across diverse Alpine settings, the dataset enables comparative analyses, model benchmarking, and robust assessments of hydrogeological processes and climate change impacts.
- Preprint
(12638 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on essd-2026-707', Pierre-Yves Jeannin, 01 Oct 2026
- AC1: 'Reply on RC1', Felix Joger, 13 Oct 2026
-
CC1: 'Comment on essd-2026-707', Ruben Carl Kubina, 08 Oct 2026
Dear authors,
Thank you for the excellent dataset and for your work.
I have a few comments regarding the dataset and the preprint.
- The attribute `ALPAKAS_ID` is spelled differently in the files `ALPAKAS_station_meta.csv` and `catchment_approx.geojson`.
- The attribute `catch_expert_shp` is set to “false” for every spring, even though you provide the delineations of 151 expert delineated catchments.
- In the attribute `area_catch_expert` there are many zeros that should likely be replaced with “NA”.
- I found one discharge time series of a daily station in the hourly stations folder (FR-0001) with wrong labels of `temp_res`. For this measuring station, the hourly and daily values should be identical, but they differ from one another. Perhaps you could double-check whether the hourly monitoring stations actually have an hourly resolution.
- In appendix A, you could describe in more detail which variables were used for the water balance to estimate the recharge area. In your code you use the total precipitation, the total evaporation, surface runoff, snow depth water equivalent, and the soil water. However, it is not clear to me whether all of the variables are actually used for the provided approximate catchments.
Citation: https://doi.org/10.5194/essd-2026-707-CC1 -
AC2: 'Reply on CC1', Felix Joger, 13 Oct 2026
Dear Ruben Carl Kubina,
Thank you for your detailed comments on the dataset and code. We find them very valuable and appreciate your contribution. We will address the points you raised as follows:
- Thank you for pointing this out. We will correct this in the GEOJSON files to consistently use “ALPAKAS_ID”.
- We will correct this metadata in the dataset.
- Thank you for highlighting this issue. It affects catchment delineations that could not be published due to licensing restrictions but for which we provide catchment attributes. We will correct this so that the actual catchment areas are displayed instead of zero values for those sites.
- Thank you for making us aware of this error. This time series is actually only available at daily resolution. The observations were incorrectly assigned a time of 00:00 and subsequently aggregated using time-weighted averaging, resulting in daily values that differed from the original observations. We will correct this by providing the original daily observations and updated static hydrological attributes in a revised dataset version on Zenodo. We will also update the corresponding metadata in the data and code repositories. In addition, we will correct the number of hourly time series to 182 in the revised manuscript.
- We agree that this could be made clearer and will specify in the appendix of the revised manuscript that the water balance estimate accounts for precipitation, evapotranspiration, and changes in soil water and snow storage. Please note that surface runoff is not included and is therefore disabled in the corresponding code. We will make this more explicit in the code.
Citation: https://doi.org/10.5194/essd-2026-707-AC2
-
RC2: 'Comment on essd-2026-707', Klaus Hinsby, 15 Oct 2026
Congratulations on completing this huge work, and thank you for providing access to and presenting this valuable dataset, which I highly recommend publishing as is, with only a few suggestions to consider:
1) It seems that all the countries covered in the ALPAKAS study have springs listed in the MIKAS-project of IAH: https://mikasproject.org/, which I know that at least one of the authors of the paper has contributed to. I suggest referring to the MIKAS project (e.g., in the hydrogeology section) and either presenting a table or a map listing which MIKAS springs are included among the ALPAKAS springs; this way, readers can link to the MIKAS project to get more detailed information on and impressions of selected springs.
2) In addition, I suggest mapping the investigated springs onto the IHME 1500 hydrogeological map as background and present it in a figure in the hydrogeology section.
3) The possibility of including the spring locations and providing FAIR data access to the compiled data via the map viewer of the European Geological Data Infrastructure (EGDI) should be explored. In this way, the spring data can be plotted on relevant (hydro)geological maps (Like IHME 1500) and compared to groundwater levels, recharge and quality data across Europe, available via the upcoming European Groundwater Monitoring Database etc. – this would enhance the value of the data for more integrated analysis of the evolution of groundwater quantity and quality together with other EGDI data and groundwater quantitative and chemical status assessments according to EU water policies.
4) Please write a small section on the availability of quality data from the springs; even if none or only little data exist, it is of interest to know if such data is available. I realise this is not the objective of the paper, but it would be helpful to know whether, e.g., monitoring of electrical conductivity or any pollutant data, are available somewhere. If quality data exist, information about where to access such data would be helpful for researchers investigating spring impacts on ecosystems or drinking water etc., e.g. in an appendix.Citation: https://doi.org/10.5194/essd-2026-707-RC2 -
AC3: 'Reply on RC2', Felix Joger, 23 Oct 2026
Dear Klaus Hinsby,
Thank you for your helpful comments and suggestions. We will address the points you raised as follows:
- We will add the Boolean attribute ‘MIKAS_listed’ to the ALPAKAS metadata file in the updated dataset and code repository, indicating whether the spring is listed among the globally important karst springs of the MIKAS project. We will also update the corresponding attribute description in Table D1 of the manuscript and refer to the MIKAS project in the main text.
- We considered adding a map showing the spring locations overlaid on IHME1500 but believe that, given its relatively coarse scale and generalized hydrogeological information, it would provide limited additional value for our study focused on karst systems. To better characterize individual karst systems, we derived hydrological indices directly from the preprocessed hydrographs and present their spatial distribution in Figure 4. The spring locations are also shown alongside the distribution of karst aquifers based on MEDKAM (Figure 1) and the structural tectonic units of the Alpine region (Figure 2). However, following your suggestion in comment 3, we are exploring the integration of the ALPAKAS spring locations into the EGDI platform, which would allow users to visualize them alongside other relevant (hydro)geological information, including IHME1500.
- We agree that integrating the ALPAKAS spring locations into the EGDI platform would improve data accessibility and make it easier to visualize and compare the data with other (hydro)geological spatial information. We have therefore contacted the EGDI team and requested to include the ALPAKAS spring locations as a point layer to the platform. As the integration process may take some time, we will continue working towards this contribution beyond the current manuscript revision.
- Thank you for this valuable suggestion. We agree that the availability of complementary water-quality data and their potential future integration into ALPAKAS should be discussed, particularly given its extensible design and the increasing inclusion of water-quality observations in other large-sample hydrological datasets (e.g., CAMELS-Chem, CAMELS-CH-Chem, QUADICA v2, and Caravan-Qual). We will therefore add a new appendix section (Appendix C) and refer to it in the main text. In this section, we will briefly outline the availability of water temperature and electrical conductivity time series, which can be obtained from various monitoring networks that also provided discharge data for ALPAKAS. Incorporating such observations could be a first step towards extending ALPAKAS with water-quality data in the future. We will also note that other chemical parameters, such as nutrients and pollutants, are less consistently available and typically measured through periodic sampling. We will refer to complementary datasets containing water-quality observations from karst springs and springs in the Alpine region. Finally, we will highlight the potential value of such data for better understanding hydrological and hydrochemical processes and assessing the vulnerability of karst systems.
Citation: https://doi.org/10.5194/essd-2026-707-AC3
-
AC3: 'Reply on RC2', Felix Joger, 23 Oct 2026
Data sets
ALPAKAS: ALPine hydrographs and catchment Attributes for KArst Springs Joger et al. https://doi.org/10.5281/zenodo.19329694
Model code and software
ALPAKAS – Code Repository Joger and Thinius https://github.com/FelixJoger/ALPAKAS
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 431 | 170 | 101 | 702 | 90 | 95 |
- HTML: 431
- PDF: 170
- XML: 101
- Total: 702
- BibTeX: 90
- EndNote: 95
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Nice effort! Thank you for this work and I believe this is a really interesting basis for future work.
I only have two concerns (beside a series a "little" remarks in the pdf):
1) In the data gathering, a strong effort was dedicated to provide parameters for catchment areas, which are mostly very uncertain. Furthermore, key-parameters controlling karst spring discharge in alpine context are not considered at all: I mean here snow cover and evapotranspiration. Therefore, I don't really understand why so much effort was dedicated to details on uncertain catchment areas and none on snow and evapotranspiration. This should be explained somehow in chapters 6 and 7, and possibly in chapter 1 as well. It could be mentioned that further parameters should be added in the future.
2) The overall uncertainty of the discharge data is poorly discussed (chapter 6). Two types of uncertainties should be discussed:
a) uncertainty about measuring the complete discharge of one karst hydrogeological system (how much water of the hydrogeological system does flow around the gaging station?), and
b) What is the uncertainty on the absolute values of the provided measurements? Are provided values measured within 10, 50, 100 or more % of the real value? Is this uncertainty the same for all values or larger for low- and high-water conditions? How significant are the differences?
This uncertainty assessment should be more developed in chapter 6, and given as a track to follow for future improvement of the data characterization.
You can find more details on these two point in Jeannin, Malard & Sinreich 2026.
As the gathered dataset is expected to be used for deep-learning techniques any bias on the data can become a major issue. I really believe that deep-learning modelling would be strongly improved by considering the two concerns above.
Therefore I recommend to accept the paper with minor revisions as it is a good paper. Revisions must include a few sentences on the two concerns mentioned above.