the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Summer subsurface temperature variability in the percolation zone of southwest Greenland: high resolution observations of the top meter of firn
Abstract. As surface melt affects larger areas of the Greenland Ice Sheet, quantifying the energetic processes governing the near-surface firn becomes increasingly important. This work characterizes subsurface temperature and its spatiotemporal variability in the upper meter of firn in the percolation zone in southwest Greenland from two months of observations in summer 2024. We provide novel methods for identifying the snow surface height from high resolution (2 cm and 15 minute) temperature string measurements and further correct the observations for apparent biases from solar heating. Using these observations, we identify several thermodynamic layers relative to the surface. The rapid-response layer is the upper few centimeters of firn or snow where subsurface temperature is highly correlated (>0.9) with skin temperature due to coupling with the atmosphere and absorption of incoming solar radiation. In the diurnally-responsive layer, temperature still responds to atmospheric variability with large positive and negative vertical and temporal temperature gradients, down to approximately 35 cm below the surface. Below the diurnally-responsive layer, the firn response to seasonal warming becomes decoupled from diurnal- and synoptic-scale atmospheric variability with depth; beneath 65 cm below the surface, correlations are less than 0.1 between subsurface temperature and skin temperature. While conduction slowly transports energy below the diurnally-responsive layer, surface melt and the advection of meltwater or latent heat can move relatively large amounts of energy that cause complex temperature gradients. Our results highlight both the value of high-resolution observations for understanding energy transfer in the near-surface firn and the need for additional observations.
- Preprint
(26785 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-1842', Anonymous Referee #1, 07 Jun 2026
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2026/egusphere-2026-1842/egusphere-2026-1842-RC1-supplement.pdfCitation: https://doi.org/
10.5194/egusphere-2026-1842-RC1 -
RC2: 'Comment on egusphere-2026-1842', Anonymous Referee #2, 12 Jun 2026
The manuscript "Summer subsurface temperature variability in the percolation zone of southwest Greenland: high resolution observations of the top meter of firn" by Sledd et al. describes the installation of a thermistor string in the percolation zone of Southwest Greenland. Advanced data processing is described, to correct for the impact of solar radiation on the thermistor string, in order to be able to reliably analyze the near-surface firn (uppermost 1m). The processing of the data is described in good detail, and seems robust, even though I have little experience with thermistor string measurements to fully judge this. The manuscript presents interesting and novel insights and analysis in the thermal behavior of the near-surface firn layer. Having said that, the manuscript is somewhat long, with some repetitive, or needlessly extensive explanations, and if the authors would simplify and streamline the manuscript a bit more, this would help with accessibility for readers. At a scientific level, I also think that the manuscript would mostly benefit from some improvements in the Results section, as I'll point out below.
Particularly when reading the Results section, it sometimes felt that the only emphasis was on internal conductive fluxes and shortwave absorption, without considering the impact of the full surface energy balance, that includes the turbulent fluxes, and longwave absorption.
Two examples:
L340: "At the surface, averaged over the full average diurnal cycle, the vertical temperature gradient is always negative (green in Fig. 9c and black line in Fig. 9d), which implies an upward conductive flux at the surface even during peak solar insolation" I don't follow the logic. If the average is negative, this does not necessarily exclude the possibility that during part of the averaging period, the conductive flux was 0, or even downward? Are there not any phases, for example during cloudy weather (thus when SWA is reduced), and a positive surface energy balance gives a downward conductive flux, even through a brief period of time?
L344: "This average negative vertical gradient at the surface is perhaps unexpected given that there should be net energy into the surface, i.e., a positive gradient, to warm it and eventually cause melt during summer" This sentence seemingly ignores the full surface energy balance. The surface melt is not governed solely by the heat flux from below.
Also, I think that the authors could try to better discuss the impact of water flow on the thermal structure of the firn. For example, L350: "Below this depth ∆T/∆z can increase to move the energy further down into the firn." Here, it should not be overlooked that when part of the snow reaches melting point and there is sufficient meltwater production, the downward percolation of meltwater is a much more efficient process to heat deeper firn, than the conductive fluxes governed by the temperature gradient. And in L351: "When the surface is no longer melting, the surface will cool first and cause upward conduction" When the surface is no longer melting, first the meltwater in this layer must freeze, before temperature gradients will become non-zero. And in fact, the statement is not generally true, as there are also phases, particularly early in the melt season when the firn is still cold, where refreezing of the surface meltwater is also driven by heat conduction downward, heating up deeper firn. So it feels like in this discussion, some nuanced description regarding the role of the full energy balance, and the role of water flow, is missing.
I have the same issue here:
L273: The authors write: "This horizontal advection, as opposed to channeling down the temperature string itself, is consistent with ice layers seen when removing the instrument at the end of summer (Fig. D2). Because we believe this meltwater was caused by the instrument support structure, we do not think it is representative of the magnitude of melt in the surrounding area."
L367: The authors write: "... when there was instrument-induced meltwater and horizontal advection and channeling to lower depths. ... Although this event had real impacts on the subsurface temperature, its magnitude and timing likely do not represent typical conditions or relationships between the subsurface and atmosphere."
While it's good to be critical on the impact of the thermistor string on melt and meltwater percolation processes, preferential flow is known to occur, and it typically has such effects on the temperature structure of the firn. See for example Fig. 8 in Humphrey, N. F., J. T. Harper, and W. T. Pfeffer (2012), Thermal tracking of meltwater retention in Greenland’s accumulation area, J. Geophys. Res., 117, F01010, doi:10.1029/2011JF002083, or Fig. 6c in Wever, N., Würzer, S., Fierz, C., and Lehning, M.: Simulating ice layer formation under the presence of preferential flow in layered snowpacks, The Cryosphere, 10, 2731–2744, https://doi.org/10.5194/tc-10-2731-2016, 2016. Ice layer formation is common in Greenland, as discussed in The Firn Symposium team. Firn on ice sheets. Nat Rev Earth Environ 5, 79–99 (2024). https://doi.org/10.1038/s43017-023-00507-9 and references therein. In my opinion, the impact of the various water transport mechanisms needs to be discussed in more detail in relation to the thermistor string measurements. While undoubtedly there was some influence from the thermistor string, I would not be surprised if the observed process of the ice layer formation did in fact occur. Or was the ice layer really only constrained to the measurement site?
Finally, for Section 3.3, the authors assume a linear trend in density to model the subsurface conductive fluxes, in order to interpolate from the initial and final density profile taken in the field. Given that the changes in density are likely not linear, because of viscosity changes due to temperature, enhanced compaction upon first wetting (as for example: https://scispace.com/pdf/the-first-wetting-of-snow-micro-structural-hardness-2islaa1vn8.pdf), and refreezing of percolating meltwater, what impact could this have on the calculated time series shown in Fig. 10?
Minor comments:
L93: "To evaluate the near-surface firn, the temperature string data must be processed and corrected" Somewhat odd phrasing. To evaluate what of the near-surface firn?
L129 and L275: "we believe": In my opinion, "believing" is not part of science. Please provide the actual arguments why you think something is the case.
Fig. 2: The sentence "Shading the temperature string gives information ... and the solar bias correction." should not be part of the figure caption, but of the main text.
Fig. 3: "is normalized vertically" I would add: "per time step".
L240: "The first month of observations began with spring conditions that transitioned to warming and eventually melt events". I think spring conditions is not well defined (as spring conditions should include warming and eventually melt). I would not use the term spring, but more use terms that describe the actual state of the firn, by using common terminology such as: cold, warming, ripening, percolation phase, etc.
L255: "the largest SWA values, up to 1 C/cm": this implies that the color bar in Fig. 7 is narrower than the actual range of values. I would mention this in the figure caption to avoid confusion.
L257: "Although the uppermost firn warms and cools with Tskin,rad," I'm not sure I understand what is meant here. That Tskin,rad is driving the warming and cooling of the uppermost firn? This is obviously not the case, because shortwave absorption is a strong contributing factor.
L279: "The latent heat from refreezing of this meltwater impacts the observed temperature structure over the following week" I would say that either: this refers to the big disturbance around June 10/11, which was not lasting for a week, or for the generally increased firn temperature in the period that followed, which would be until June 21st (and likely the rest of the summer season), and thus much longer than a week. I cannot identify a feature that lasted during a week.
Fig. D2: I think the figure should show to the reader the ice layers, for example using arrows, for those who are unfamiliar with typical visual clues one gets from snowpits.
L290: Please explain “solid state greenhouse effect”.
Throughout: mostly the term "temperature string" is used, but on a few occasions "thermistor string". Apart from the fact that I think thermistor string is a clearer term here, at least consistent terminology should be used.
L454: "According to Pfeffer and Humphrey (1996), apparent heat sinks at depth can occur when the actual thermal conductivity is lower than what is assumed, i.e., the calculated conductive flux leaving a layer is larger than the actual flux, making it appear that there is net energy lost": it is not clear why this is a one-way process. I.e., is it not also possible that apparent heat sources at depth can occur when the thermal conductivity is overestimated?
L512: "While it is clear observations with high vertical and temporal resolution are necessary for ..., it remains to be seen if such processes are necessary for modeling the seasonal evolution of ice sheets." Why would it not be important to simulate the seasonal evolution of ice sheets? The sentence feels a bit out of place, since no arguments are provided why this would be the case. That makes the sentence a bit irrelevant, I think.
Citation: https://doi.org/10.5194/egusphere-2026-1842-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 270 | 157 | 23 | 450 | 37 | 38 |
- HTML: 270
- PDF: 157
- XML: 23
- Total: 450
- BibTeX: 37
- EndNote: 38
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1