We thank the two reviewers for their comments and provide hereby our responses on how we will address the raised issues. Our answers are highlighted in italics.

RC1: Citation: https://doi.org/10.5194/essd-2025-134-RC1

We thank RC1 a lot for the constructive comments.

Review of Thermo-hydrological observatory in a permafrost river valley landscape in Syrdakh, Central Yakutia by Pohl et al. for Earth System Science Data

General overview:

This is an exhaustive dataset of a variety of commonly used measurements to characterize permafrost in an area where it is changing rapidly. Overall, it is well written and the explanations are ample to be able to repeat the measurements. In a few locations I suggest further information/clarification. This paper will be of interest to other permafrost scientists but also provides a strong example of how to design a long-term field experiment and name and organize a disparate dataset. The dataset is unique, useful, and complete. The abstract and introduction use carbon as a main motivation to measure/study permafros.t However, I would argue that other biogeochemical parameters, hydrologic changes, and of particular interest infrastructure (linear and horizontal) are also affected by permafrost degradation. Many of the measurements provided could be used to instrument buildings, bridges, airfields, or roads to identify change. Mentioning these additional reasons to study permafrost would strengthen the impact and relevance of the work.

We thank the reviewer for this assessment, and we will add relevant references.

The tense bounces around a lot. For example, around line 254 “were conducted”, at 257: “is provided”, 262: “are based”, 267: “was used” and throughout. Maybe this works to bounce around like this but it is worth making sure it is as consistent as possible.

We will thoroughly check the tenses throughout the manuscript and adjust them where relevant.

Comments keyed to the text:

2: permafrost-affected

The meaning is “regions with permafrost that are affected by permafrost thaw”; We will adjust the sentence to make this clear.

3: use a different word so as to not use “ubiquitous” twice in a row. Or remove second instance- not needed

OK.

7: provide total depth here

OK.

10: what parameters?

The idea here was to provide a concise overview of measured variables and parameters and not list everything. We would change the sentence to “Air, water, and ground temperatures, as well as physio-chemical river water and soil properties of the surrounding environment were measured as time series or single tests during annual field campaigns.”

12: GIS database

OK.

37: define “thermokarst lakes” here- what are they and how do they form? What do they represent?

We will adjust the sentence to “[…] received rather limited attention are small rivers and creeks that connect or drain newly developing lakes, which form because of permafrost thaw (e.g. Liu et al., 2022; Tananaev and Lotsari, 2022).”

42: put talik definition in parentheses instead of using the “i.e.”

OK.

49: is “Yedoma Permafrost Domain” a real name?

I recommend providing a sentence identifying/defining yedoma as high ice content, high carbon and organic matter content permafrost comprised of fine-grained soils and that due to these characteristics state that yedoma is particularly susceptible to thermokarst formation. This is currently mostly provided at 66-69 but is better to move here.

It is a coined term in the referenced article but not in general. We will adjust the writing and move the relevant text passage to this earlier instance.

Provide general latitude and longitude here.

We will reference Figure 1 at this point in the text to provide the geographical reference; outlining the Yedoma region is not straight-forward due to its patchy and wide-spread occurrence. We will reformulate the sentence and add a reference to the relevant work on Yedoma distribution.

73: you defined taliks earlier

We will merge this.

80: reference for these protocols?

We extended the monitoring network following no specific protocol. After the two main cross sections (near and far away from the lake) were installed, the focus was to capture more variability along the river. We will add a sentence on this.

86: in addition, the focus

OK.

88: was installed

OK.

109: and creating

OK.

110: depths and velocities

OK.

Table 1. This is a great list of measurements! Font is extremely small though. Perhaps make it landscape to provide room for a larger font?

We will make it landscape format and allow for a table break.

Is there a day/date associated with the thaw depth measurements to discern between seasonal thaw and active layer?

Yes, each measurement has a date associated with it. This information is contained in the actual data files “ground/aclt/”. We are adding this information in the text.

Maybe break the active layer/thaw dataset into two since it is important to know whether it is a true active layer or just seasonal thaw at a given moment tin time. At the least the “aclt” label is misleading if it means “active layer thickness.”

All the measurements conducted in the field should first be seen as thaw depths. However, as all these measurements were conducted at the end of the thawing period (mid to late September), these depths can be assumed to be close to the actual active layer depths. This is supported by the temperature time series that show the cooling onset throughout all boreholes around August to September. How long and how much deeper the thaw is occurring during the cooling of the ground has not been assessed. We will rename the variable to thaw depth (td) all together to avoid this variable name to be misleading.

Figure 1. I recommend splitting this into a few Figures either here or in an appendix.

In accordance with our response to RC2 on this issue, we will split the figure to provide a better overview of the landscape setting in Figure 1 first. A second figure will then provide a more detailed view on the two instrumented field sites.

Red labels in B are unreadable

We will remove all labels from the overview figure and add better visible labels in the detailed views of Site 1 and Site 2.

Maybe keep A and B as is (Figure 1) but without larger labels on B and then provide C and D as Figures 2 and 3.

In line with Reviewer 2, we will split Figure 1 into two parts; one showing the overview and landscape features, and a second figure that shows the two main sites along the small river that connects the thermokarst lakes.

155-161: any information on how these installations were done? Maybe a schematic cross section? Or a description? For the HOBO loggers the borehole diameter and construction?

We will add additional information on the installation.

Figure 3. State that the depths of these thermistors are provided as, for example “300” (which I assume means 300 cm).

We will add this information in the figure caption.

203-207: are there any “true” active layer depths in the area to compare to?

To our knowledge there are no “true” active layer depths measurements available at Syrdakh or in the vicinity. Also in response to RC 2, we will rename active layer to thaw depth throughout the manuscript and adjust the file names as well to do this justice. We will check on other studies reporting thaw depth measurements in the wider area and this if feasible

Figure 5. Also provide some of these measurements over time for given locations to look for trends? Maybe the different ties/dates in September make this not worth doing? Would be good to at least provide the plot and some discussion.

We understand the ESSD format as strictly data revolving and therefore decided to not run any analysis that is not related to data presentation or quality.

Or a separate plot of just “true” active layer values (presumably you have this for at least 2017 and 2028, maybe other years?).

Without a continuous observation technique identifying the thaw depth, there is no chance to report the true active layer depth. Because all field campaigns were conducted in early autumn, we believe that the thaw depth measurements are close to the true active layer depths. But this remains unverified. We will try to make this even more clear in the text, where we have already mentioned this. By entirely renaming the variable to “thaw depth” and changing the variable abbreviation accordingly, we hope to now clearly prevent this misunderstanding.

235-242: This seems to be in the wrong place. It is not related to the ERT.

Yes, this will be moved up to the correct section.

307: the true active layer or seasonable thaw at that time?

As outlined above, we will remove the ambiguity and refer to all thaw depths now really as thaw depths.

361: “do not derive significantly”- not sure what this means. Do not drift?

Yes, this is a spelling mistake. We will change this.

366: “do not derive significantly”- not sure what this means. Do not drift?

Yes, this is a spelling mistake. We will change this.

Figure 14. Provide “north and south” or whatever directions work for this cross section (i.e. the -1000 and 1000).

We will add “left bank” and “right bank” as the sections might have slightly different orientations.

433: “do not derive significantly”- not sure what this means. Do not drift?

Yes, this is a spelling mistake. We will change this.

Appendices provide wonderful photos of field site activities and measurements.

We will use an oblique aerial view photo in Figure 1 to give a better impression about the landscape.

RC2: Citation: https://doi.org/10.5194/essd-2025-134-RC2

We thank RC2 a lot for the constructive comments.

The manuscript describes a rich dataset from multiple environmental parameters collected in a study area located between two lakes connected by a stream in a permafrost area in Central Yakutia. Despite the small size of the study area, its geomorphological and hydrological setting, together the rich set of variables measured makes this data set of high scientific value for studying the dynamics of permafrost environments.

The manuscript is well-written and the dataset is presented in a generally clear way. It is associated to a rich dataset, which has several gaps and also some problems due to changes in instruments and techniques, with large uncertainty in some variables. However, given the harsh environmental conditions and timespan of the measures, this could be expected, and most key data is highly relevant. The authors decided to make all data available and this is of high value. 

Overall, I missed a better description of the geographical setting of the study site. Actually, it was only after reading the whole document and when inspecting the Zenodo dataset, that I found a clear picture that encompasses the landscape setting. It is an oblique aerial picture that you should include in the body of the manuscript as support for the site description. An inspection of Bing virtual earth imagery of the area shows the remarkable landscape setting with thermokarst lakes, but this is almost entirely missed in the manuscript. A better description of the geographical framework accompanied by illustrations would increase the potential interest and usage of the dataset.

We will make this clearer. Inside an adjusted paragraph about the site description, we will reference an overhauled Figure 1 that shows a more general overall setting. It will include an oblique aerial UAV photo, as well as a Google Earth overview image, covering the two lakes and the entirety of the Syrdakh valley that connects the two thermokarst lakes. This combination should provide a better basis to understand the landscape setting.

In what concerns to the dataset in Zenodo, one important issue is the lack of the QGIS project file that should allow to open all geographical data. This is important and needs to be made available for review and for final acceptance of the manuscript.

We are sorry for this oversight and will of course upload the project file.

Besides these more general comments, below you may find a list of edits and some comments to be addressed by the authors.

Title: "Thermo-hydrological observatory in a permafrost river valley landscape in Syrdakh, Central Yakutia". Consider mentioning Yedoma in the title and add the period of observations.

We propose to change the title to "Thermo-hydrological river valley observatory in Yedoma permafrost from 2012 through 2022 in Syrdakh, Central Yakutia".

Line 78. Describe the thermokarst lakes, as their characteristics are relevant to the hydrological dynamics of the stream. A better description of the catchment is also needed.

We will do so.

Line 95. The name Quantum GIS was abandoned and the software is named QGIS.

We will adjust this.

Line 96. "at various depth" should be "at various depths".

OK.

line 101. 3.2 "Data location" should better read as "Location of observations" or "Location of observation points". Consider improving this subsection title.

OK.

Line 102-115. Section 3.2. The problems with the position of the data collection points are not clearly explained. Make it clearer as a correct understanding of the positioning is essential. Consider presenting the non-corrected (or original) coordinates, as well as the corrected geographical coordinates. As it is, it is not straightforward to understand the problems.

We will rework the section to make this part clearer. We decided that the most important aspect would be to have the instruments placed consistently with the landscape feature they were installed at. That means that their relative position would be the most important thing, e.g. a river depth sensor should be in the middle of the river, or a sensor in a depression is really in a depression, whereas the absolute position would be not important.

There are large uncertainties in the measured coordinates regarding accuracy (differential GPS + handheld GPS) and precision (handheld GPS), and due to the recording of positions in multiple surveys (new baselines for the differential GPS). As a result, there are inconsistencies even of a same point between different years. Therefore (this includes the reason mentioned in the first paragraph of this question as well), we do not think it makes sense to provide all of these coordinates. 

We will, however, provide an elaborated description as of why and how we adjusted the measurement points.

Line 123: Rather than presenting this as an equation, it should be referenced as the file name structure.

We will not present this as an equation but move it to an in-text explanation.

Line 128: Modify order: L0, L1, L2

OK.

Line 134. It is not an equation.

OK.

Line 134. Metadata and not meta-data

OK.

Lines 144-145. "Monotonous temperature shifts": does that this mean monotonic? Clarify the sentence. Do you mean isolated peaks? always n=1?

We will rewrite the sentence. This was meant to say indeed monotonic, and n=1 was supposed to highlight that this occurred only once. We will reformulate this paragraph to make it clear.

Line 155. Indicate and explain the time-intervals of the measurements, as in Table 1, which is not very clear, there are measurements with intervals from 1h to 1 year.

We will make this more clear and add this information.

Line 160: Indicate all measurement depths in the thermistor chains as well as the installation setup (boreholes, or other; type of casing, casing diameter, infill medium)

We will add more information on the installation setup and installation depths.

Lines 198-199: Rephrase in order to clarify: "The calculation of thermal properties of the KD2 Pro device might have high uncertainties. This is because significant differences in some cases between the also utilized Anritsu thermometer are apparent".

We will rephrase this sentence for clarity.

Lines 214-217: The subsection is titled boreholes and piezometer tubes, but it does not contain information about the borehole setting nor piezometer tubes. Check the text and add information about these.

We will do so.

Line 220: "Upper thawed layer depths" should rather be "when thaw propagation was maximum"

We will correct this throughout the manuscript and also change the variable name (currently “acld”) to reflect that this variable is always representing thaw depth at measurement time.

Line 229: 71.5 m

OK.

Line 276: "Deviations from the reporting classes exist." Please clarify what this means. The reader figures that the Malvern is used after 2017, but still the "deviations" part needs to be clarified.

We will clarify this.

Line 324 - Correct typo "unceertainty".

OK.

Line 331 - Indicate if a solar radiation shield was used and type. From a photo in the dataset I see a logger attached to a tree. In what way does it affect temperatures? What is the aspect? Does it stay shadowed all year? These need to be addressed.

We agree, and we will add the relevant information.

Lines 331-332 - Why do you make reference to section 4.3.1?

This was done to reference the instrument description of the TB instruments. We will clarify this in the text.

Line 333 - Indicate the distance of the Yakutsk meteorological station.

OK.

lines 336-337 - Rephrase. The reference to "reasonable representation of air temperatures" is subjective. You may mention that the high coefficient of determination supports the good quality of the data and the regional climate signal's influence on local conditions at the study site.

Thank you, we will do so.

Line 378 - Correct "a coefficient"

OK.

Line 437: Indicate if it is WGS84 UTM Zone 52N or other datum.

We will clarify this.

Line 441: You mention the use of GCPs in 2021, but please indicate the accuracy. It is also important to know the image overlap in the flight, as well as flight height, if available. In the following subsection you mention a 0.5m absolute accuracy. Is this horizontal? What about vertical? How was the correction made, as 0.5m is really a large positional error.

We will add additional information on the processing of the imagery for the DSM generation.

Line 448. You mention Leica Viva Uno 10 (GPS+GNSS). Do you mean GPS+GLONASS?

Yes, we will change this.

Line 453. See comment for lines 110-115.

We will elaborate on this as well and make it consistent with the description in the earlier section.

Figures and Tables

Figure 1. A map is needed showing the framework of the site, especially showing the two lakes and the general landscape features.

Figure 1 is split into two figures (also following RC1) to show first the framework and general setting. This will give a better impression about the thermokarst lakes and a better understanding of the landscape in general. The second figure will show a more detailed view of either of the two main sites (Site 1 and Site 2).

Figure 3 - The depths from the sensor names do not correspond to the boundaries in the Y axis.

This is not clear to us. We checked each sensor name and the depth of a sensor indicated in the file name (as shown in the figure legends); we find that these depth descriptions are matching the depth range indicated by the label on the y-axis in all cases. RC1 mentioned this was not very clear. We will elaborate on the figure caption.

Table 1 is very complete but the font size is too small for reading. Consider improving it. Measuring intervals are not clearly understandable from the table. They should be readable according to instrument types horizontally.

We will convert the table into landscape format and allow for multipage splitting to increase font sizes. Measuring intervals were supposed to show each possible measuring interval; we change this to the range of measuring intervals, e.g. “(2 to 8) hours”.

Zenodo dataset

- The QGIS project file syrdakh_gis_2025-06-10.qgz is not in the dataset and needs to be included.

We are sorry for this oversight and will of course add the project file to the repository.

- DEM: Actually it is a DSM and not a DEM as correctly described in the manuscript, but the file (and those mentioning a DEM) must be named accordingly. I also recommend masking the DSM using the boundaries of the Orthomosaic, as there are large parts of the DEM that are just noise and should be removed. If the data is to be used by modelers, then the DEM should include only the area with good data (despite the other severe limitations with slope that are mentioned in the manuscript). A good way to make this data useful, especially given the poor GCPs in the terrain, would be to make all aerial photos available in the Zenodo dataset. This would allow for future reprocessing.

We will rename the DEM to DSM and adjust all instances where DEM was mentioned. Due to publishing of some of the raw imagery as single scenes in in another journal, we prefer to not make all the UAV imagery available but will include a statement in the data availability section. We will add that these specific raw data, which themselves are not part of the here presented dataset, can be obtained through one of the co-authors.