The paper describes data from two adjacent sites in the forest-tundra ecotone in Northern Quebec. 

With one station in the tundra landscape and one in the forest this data set covers an important transition zone. 

The data will be of great interest for validation and improving of (climate) models. 

Specific Comments

Abstract

I would like to read a sentence here about your gap-filling practices

2. Site Description

line 80: The presence of lithalsas are not the reason for the permafrost, but the indicator why we know/think there is permafrost. Please rephrase

3. Climate Data

Table 1:

could you make the rows more distinguishable? Shade every second row or something?

I suggest to use negative numbers for soil depth to make it more clear which numbers are height over surface and which are depth below surface

line 212: This sentence suggest you would allow precipitations value between -30 and 0 mm/h? There are now negative precipitation values in the data. 

7. Soil Data

Figure 15/16: could you make the accuracy of the tic marks and the grid the same for these two images? The resolution of your tic marks and the white space and the axis limits are often not the same for the pair of TUNDRA/FOREST images.. 

line 422: Why did you choose to describe the soil properties after the measurements of temperature and soil moisture? I would suggest to change that, as it makes the interpretation of the soil moisture plots more intuitive

Technical Comments

line 24: "The data is available" instead of "The data are available"

line 345: delete "the" in ".. with these heights, as the there is.."

The data in this publication provide easy-to-access multi-year, high-resolution in situ records of environmental data from remote locations in the southern Arctic forest-tundra ecotone.  These data are difficult to obtain and, as validation and calibration inputs for satellite and modelling studies of climate-related changes in the far north, may serve as important contributors to future understanding of ecological changes in this dynamic transitional region.  Missing data are generally filled in with replacement data from nearby comparable instruments, regression on the adjacent data, or other model outputs (such as ERA5), so that uninterrupted time series are available.  Quality flags are applied to some variables.

The data publication is comprehensive, to introduce, explain, and visually represent this complicated and extensive data set, and provides sufficient references for further reading.  The data set is valuable, apparently unique, and useful.

Detailed edits/questions to the manuscript:

Lines 104-105:  Air temp and relative humidity - missing data (small gaps) provided by “another sensor” nearby at 10 m height - which instrument?  Please describe/give details?

Table 1 - Should “Air temperature” also include “and humidity” as in Table 2?  “themistor” needs an “r” and “thermcouples” needs an “o”.  Second “Soil temperature and volumetric“ - should it have “water content” added to the end, like the one above?

Table 2 - what are the dates for Snow Thermal Conductivity (the table currently says “201–2021”)?

Lines 272-274:  “Variations between sites were smallest for lichen, increased for birch and spruce, and were highest for low grassy vegetation, Variations in birch and spruce are probably mostly due to differences in the leaf area index and in the amount of woody vegetation present.”  This is two sentences and should have a period instead of a comma at “vegetation, Variations”

Also, if I understand it correctly, Table 4 suggests that the variation is largest for spruce, not for grass, as is stated.  Please review and revise as needed.

Lines 291-292:  “To align the coordinate system with the surface, we have chosen to apply a double rotation.”  Please provide a reference or brief explanation, as the procedure is not clear, as written.

Figure 8 - Instead of 4 months, could the x-axis be given in 6 month periods, or seasonal (3 months), so that it is easier to visualize the annual pattern?

Line 345:  Remove “the” before “…there is no consistent line…”

Figure 11 - Please check the colours of the legend, as the red line is always between the green and blue lines, which would indicate that the 14 cm depth is always warmer than the 4 cm sensor.  This is different than the temperature patterns in Figure 10 and is confusing.  Are the colours in the legend correct, or are the colours of 4 and 14 cm mislabelled in the legend?

Section 6.4 - Do you intend for all dates shown in Figure 14 to be described/accounted for in the text?  Please check the dates in the text description against the dates of the actual measurements, as it looks like 2012 is missing from the text.

Figure 14 - The legend is covering over some of the data in both plots, but especially the Density plot.  Please move the legend so that all the data can be seen.

Figure 16 - Should “volume” be removed after “(SWC)” in the Figure description text?

Figure 18 - Please write dates with the same format as in Figure 14.

Line 453 - “…which showed poor performance when simulating Arctic snowpack properties.”  Perhaps add “have previously” before “showed” and provide a reference(s)?

Lines 466-467 - I assume these lines will be removed upon publication?  Along with “(dataset in review)” that is currently written twice in the header of each data file?

Data files:

In the data files, what is the time zone of the time stamp?  Is it local time?  If so, which time zone is it?  This should be specified in each data file and/or in the manuscript.

Are the files tab-delimited?  The format of the files should be stated in the manuscript.

Please check the headers in the different data files to confirm that they are all consistent.

Umiujaq_rad_forest.tab and Umiujaq_rad_tundra.tab:  There are only 3 columns of data in some places, instead of 4 (for example, 2021-10-14T07:00 to 17:00, in both files).  Please check entire files for complete data presentation.

Why are there no quality flags in Umiujaq_rad_forest.tab (but they are there in Umiujaq_rad_tundra.tab)?

Umiujaq_rad_tundra.tab:  QF SWD should not have a decimal point (should be just one digit).

Umiujaq_precip_tundra.tab:  There are sections where there is rain (mm), but no precip (mm/hr) for many hours.  Is this possible?  For example, 2019-07-10T15:00 through 2019-07-11T19:00.  Can you explain the calculation method or the function of the field instrumentation which would allow this seeming inconsistency?

Umiujaq_snow_density.tab and Umiujaq_snow_surface_area.tab - what do locations A and B mean?  Please explain/briefly describe in the manuscript.

Please explain if snow surveys were not actually taken at midnight (all time stamps indicate 00:00).  This also applies to Umiujaq_snow_temp_conduct.tab and Umiujaq_snow_temp_conduct_st2.tab files - time stamps are all 00:00.

Umiujaq_snow_density.tab and Umiujaq_snow_surface_area.tab:  Lat/long are missing from many of the 2018 FOREST density and surface area measurements.  Please explain.

Umiujaq_snow_height.tab:  “heght” in the header (above the “Keyword(s)” line) needs an “I”

Umiujaq_snow_surface_area.tab:  “differnt” (just above “Keyword(s)”) needs an “e”

Umiujaq_snow_temp_conduct_st2.tab:  Temperatures are missing from some intervals, including 26 April 2019, 6 May 2019, and others.  Please explain.

Umiujaq_snow_temp_conduct.tab:  Temperatures are missing in some intervals, conductivity is missing in others, and both are missing in a fair number of intervals.  Why include these rows, if there are no measurement data at those times?  Is it because it is summer and there is no snow?  Why is this file made differently than Umiujaq_snow_temp_conduct_st2.tab, which seems to have removed all the summer months with no snow?

Umiujaq_temp_wind_forest.tab:  Why is there no quality flag for the Temp in temp_wind_forest, though there is one in Umiujaq_temp_wind_tundra.tab?