the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 16 Dec 2025
CAMELS-FI: hydrometeorological time series and landscape properties for 320 catchments in Finland
Abstract. Comprehensive, large sample hydrological datasets, such as CAMELS (Catchment Attributes and MEteorology for Large-sample Studies), have provided the basis for advances in many aspects of hydrological research in recent years. They can be utilised for several purposes, such as training data-driven hydrological models, comparisons between regions dominated by different types of hydrological processes and testing of general validity of hydrological theories. The value of these datasets is in combining a multitude of data sources into one, easily accessible and usable, harmonised and high-quality package. We present CAMELS-FI, an extensive hydro-meteorological dataset for 320 catchments in Finland. It combines hydrological and meteorological time series with biophysical and human influence catchment attributes in a format that enables comparisons between catchments within the dataset but also between earlier CAMELS datasets. CAMELS-FI includes a diverse set of catchments with human influence varying from near natural to heavily regulated. CAMELS-FI is available at https://doi.org/10.5281/zenodo.15853357 (Seppä et al., 2025).
- Preprint
(9193 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
CC1: 'Comment on essd-2025-578', Matteo Rosales, 23 Dec 2025
Dear authors,First and foremost, I wanted to tell you that I found your dataset to be very well designed and organised. The article is likewise very pleasant to read, including the figures.Nevertheless, I have come across a minor problem with gauges' coordinates, and I was wondering if was using your data correctly. In the catchment metadata, I found two lists of coordinates, respectively in EPSG 4326 and EPSG:3067, listed as follows :gauge_lat gauge longitude (EPSG:4326) ◦gauge_lon gauge station latitude (EPSG:4326) ◦gauge_easting ETRS-TM35FIN coordinates (EPSG:3067), easting mgauge_northing ETRS-TM35FIN coordinates (EPSG:3067), northing mHowever, it seems to me that the lat/lon coordinates do not match the easting/northing ones. Below is what I see when I load both sets of coordinates in QGIS; as you can see, the brown dots (lon/lat) greatly differ from the yellow dots (easting/northing).
Additionnaly, I have tried to snap the stations on a European river network (CCM) with the upstream areas you provide. Whilst the EPSG 3067 coordinates seem to work fine with the network,the EPSG 4326 does not. Therefore, I was thinking there might be an issue with your lon/lat coordinates. Or maybe there is a distinction between the two gauges that I did not grasp.I hope this can be useful feedback,Kind regards,Matteo ROSALESCitation: https://doi.org/10.5194/essd-2025-578-CC1 -
AC1: 'Reply on CC1', Iiro Seppä, 23 Dec 2025
Dear Matteo Rosales,
Great to hear that the dataset is of interest to you and you found the article pleasant to read!
Heartfelt thanks for noticing the issue. The geographic coordinates in metadata attributes were indeed showing wrong locations. While investigating this issue, we also noticed that the projected coordinates in metadata attributes had not been updated to match the final location of the gauges in the data/CAMELS_FI_catchment_boundaries.gpkg. The discrepancies with projected coordinates were generally minor (typically tens of meters or less, maximally about 0.5 km). The coordinates in metadata attributes have now been updated to match the locations in the geopackage, which was quality controlled more thoroughly during the development. An updated version of the dataset (1.0.3) containing the changes has been uploaded to Zenodo.
Best Regards,
Iiro Seppä & other authors
Citation: https://doi.org/10.5194/essd-2025-578-AC1
-
AC1: 'Reply on CC1', Iiro Seppä, 23 Dec 2025
-
RC1: 'Comment on essd-2025-578', Franziska Clerc-Schwarzenbach, 11 Feb 2026
Dear authors,
please find my review of this manuscript and dataset in the attachment. There is a comment for every highlighted instance in the commented manuscript. I hope it is accessible to you, otherwise, feel free to reach out.
Best wishes,
Franziska Clerc-Schwarzenbach
-
AC2: 'Reply on RC1', Iiro Seppä, 10 Mar 2026
Dear Franziska Clerc-Schwarzenbach,
many thanks for the thorough review, and especially for the suggestions on how to improve the dataset. We have revised both the manuscript and the dataset, and believe both to have been enhanced by the changes. Please find attached the replies and actions to your comments. We will submit the updated manuscript after the comments of the second referee. The link to data has not changed despite revisions, it is still https://doi.org/10.5281/zenodo.15853357 .
Kind regards,
Iiro Seppä and the other authors
-
AC2: 'Reply on RC1', Iiro Seppä, 10 Mar 2026
-
RC2: 'Comment on essd-2025-578', Anonymous Referee #2, 17 Apr 2026
Review of the manuscipt “CAMELS-FI: hydrometeorological time series and landscape properties for 320 catchments in Finland” by I. Seppä et al. at ESSD.
The manuscript and associated data are well written, well organized, and they offer a valuable contribution to hydrology. Please find my main concerns below.
Major comments
1. L 106. “A Minimum of five years (1826 days) of observations required. We considered this sufficient to calculate streamflow signatures.” Many CAMELS data sets, including the first one for the USA, require about 20 or 30 years of Q observations for inclusion. For long-term studies, WMO recommends at least 30 years of hydrometeorological observations. The reason for these stricter requirements is that hydrometeorological variables often have long-term persistence (e.g. O’Connell et al., 2022) and temporal clustering (e.g., Chagas et al., 2024; Lun et al., 2020), thus, 5 years of data may not be representative of long-term hydrological behavior. In order to include the hydrological indices of catchments with fewer than about 20 years of data, it would be interesting for the manuscript to assess whether these values are representative of long-term conditions, for example by comparing with neighboring catchments that have longer data.
2. L 211. “In addition to providing the aforementioned data, we combined the data into one convenience PET attribute from 1981, which was also used for calculating climatic signatures (see section 5.2). This was done by using snow evaporation when the snow depth of the catchment was over zero, and filling the snow-free days with FMI PET if it was available; the remainder was completed by ERA5-Land potential evaporation.” Combining the time series from two different sources can be problematic because each source is usually built from completely different data and models, particularly evapotranspiration variables which often have high uncertainties. In order to keep the combined PET time series, it would be interesting to include a more in-depth investigation of whether that is appropriate, including comparisons with other data sources, where possible, and assessing if it does not introduce step changes in PET and other spurious artifacts.
3. Section 4.1 Hydrologic time series. The manuscript could benefit from a more in-depth description of how the streamflow measurements are conducted by the data providers (SYKE and ELY), including how the measurements have changed over time (if such information is available), how the data were quality checked, and whether data quality flags are available and included. Some description is already present in Section 7, but I believe that it could be expanded. If the data providers have documentation describing how the measurements are or were conducted, it may be worth citing it in the manuscript.
4. I could not find geological indices. The manuscript mentions that hydrolithological data are not available. However, the manuscript could include other variables such as geological type, to align the data set more closely with the CAMELS data sets.
Minor comments
5. Fig. 1 and Section 5.6 (Human influence). Does the regulation referred to here indicate regulation through artificial dams? What are the criteria for defining the regulation classes (not regulated, minor and major regulation)? It is not clear in either the figure caption or the section.
6. Table 1. The difference between gauge_id and basin_id is not clear. The gauge_id is described as “catchment identifier”.
7. Table 3. How was the attribute “regulation_level” calculated?
8. L 128. “Kammonen, Luiro (1358)”. Is 1358 the gauge_id? It may be worth clarifying.
References
Chagas, V. B. P., Chaffe, P. L. B., & Blöschl, G. (2024). Drought-Rich Periods Are More Likely Than Flood-Rich Periods in Brazil. Water Resources Research, 60(10), e2023WR035851. https://doi.org/10.1029/2023WR035851
Lun, D., Fischer, S., Viglione, A., & Blöschl, G. (2020). Detecting Flood-Rich and Flood-Poor Periods in Annual Peak Discharges Across Europe. Water Resources Research, 56(7), e2019WR026575. https://doi.org/10.1029/2019WR026575
O’Connell, E., O’Donnell, G., & Koutsoyiannis, D. (2022). On the spatial scale dependence of long‐term persistence in global annual precipitation data and the Hurst Phenomenon. Water Resources Research, e2022WR033133.Citation: https://doi.org/10.5194/essd-2025-578-RC2
Data sets
CAMELS-FI Iiro Seppä et al. https://doi.org/10.5281/zenodo.15853357
Model code and software
CAMELS-FI Iiro Seppä https://github.com/iiroseppa/CAMELS-FI
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 619 | 426 | 45 | 1,090 | 38 | 29 |
- HTML: 619
- PDF: 426
- XML: 45
- Total: 1,090
- BibTeX: 38
- EndNote: 29
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1