The manuscript entitled “ICESat-2 surface elevation assessment with kinematic GPS and static GNSS near the ice divide in Greenland” by Pickell and others details a new use for the Open GNSS Research Equipment (OGRE) GNSS-IR stations—validation of ICESat-2 surface height measurements. This study first uses data from repeat kinematic GPS surveys at Summit Station, Greenland to assess the stability of ICESat-2 surface height measurements, finding a <0.01 m bias and <0.06 m precision between kinematic GPS and ICESat-2 observations. The authors then present a new validation method using GNSS-IR interferometric reflectometry to measure surface elevation coincident with ICESat-2 passes over their study site near Summit Station. Ultimately finding a good agreement between GNSS-IR and kinematic GPS observations and a bias in ICESat-2 measurements of ~0.09 m. These results indicate GNSS-IR can be deployed instead of kinematic GPS transects which are significantly more resource intensive. 

I found the manuscript to be very well written with a robust methodology and results section wherein the results were well supported by the presented data. The manuscript would benefit from an elaboration on some methodology (detailed below) and a reorganization and expansion of the discussion section. I have detailed these points below and include some additional minor comments. Overall this work is appropriate for the Cryosphere and I would therefore recommend minor revisions before publication. 

Major comments

(1) I would like some clarification on the bias corrections applied to the dataset. Section 3.3, Lines ~155. Here a bias of 2.1+/-2.9 cm is reported, citing figure A1. I have a few questions regarding figure A1, but with the main concern that this bias is not temporally consistent. Is the bias applied across the full time series as stated? It appears from Figure A1b that there is indeed a large bias in 2022, a moderate bias in 2023, and somewhat in early 2025. Why not apply a variable bias correction? Or not at all as it is not convincing that a bias correction of 2.1+/-2.9 cm would improve your observations. Regarding Figure A1: Which OGRE is this data from? Where was the stake relative to this OGRE and the study area? Also was there a camera that allowed such frequent stake readings?

(2) Figures supporting methodology (either added to manuscript or supplement) L75-78: What is the range of track depth Z_track values due to vehicular weight depressing the snow at the beginning and end of your survey and the mean value used? L78-79: How did the laser range-finder measurements compare with the mean value of Z_track measured? Were there any systematic trends in this value? E.g., increasing depression along the survey track? A figure showing the Ztrack observations and the laser range finder observations would be beneficial.

(3) I found that the discussion would benefit from some slight reorganization to make it easier to follow for the reader, in particular by separating the discussion of temporal and spatial sources of bias which are slightly intertwined (mainly the details presented in the middle to the end of paragraphs). Overall, the discussion does have a good organization by moving from those errors to external factors (e.g., blowing snow/surface roughness) then (correlated errors). But again, the authors should be careful to group like-ideas together (e.g., a discussion of the sensing footprint on line 248-249: is this an uncorrelated error or should it be moved to the spatial or surface roughness paragraphs? If it is kept in its current position the similarity of the sources of errors in the data should be made more evident. 

(4) The Manuscript would also benefit from an expanded discussion of the implications, next steps, or synthesis of the work presented here. For example, the introduction and abstract mention that kinematic surveys are much more labor and resource intensive than the OGRE station deployment, since these results demonstrate OGRES are a useful tool to assess ICESat-2 surface elevations, what are the authors recommendations moving forward? I would be great to hear their thoughts for how future campaigns aiming to assess airborne/space-borne surface elevation measurements should proceed. These topics are particularly relevant given NASA’s Snow4Flow program.

Minor comments

Figure 1 (a): I would suggest the north arrow be positioned on the top of the figure as it seems out of place near the scale bar. The legend for ICESat Traverse Route is somewhat misleading

The authors should also adjust the font size in various labels to ensure they are large enough to read. Even when the figure spans the entire page width, some labels are very small (e.g., RGT #’s, “spacecraft travel direction”, the “10 km” scale bar, Surface elevation color bar, etc).

Figure 1 (b)  What is R_5?

Figure 1 (c). This is a great cartoon of the kinematic GPS surveys. A small note is that H_R is capitalized in the figure but H_r is mentioned in the caption, update whichever to make sure symbology is consistent, also enlarge Z_track (next to the arrow) in the figure.

L85-90: Here only 24 hours of data are collected, you can expect the largest errors in positioning at the beginning and end of an observation period, and depending on the processing procedure, at day-breaks. Were longer periods of data collected and it was found that 12-hours before/after was optimal? 

Figure 2: the x marking the median is difficult to see, I suggest changing the symbology, perhaps a - that is longer than the underlying point measurements are wide would be easier to see?

Section 2.2 and 2.3: What are the CSRS reported horizontal and vertical (Zppp) errors? These should be included in these sections for both types of GNSS stations. I know this is discussed later on but is important to include here as well. You can refer readers to Section 3.3 for a more detailed discussion. 

L147-149: Here the authors state that observations from the full 24-hour period are used to determine this 1.4 cm measurement precision, my question is, are edge effects (at the beginning and start of your time series) are removed or special filtering is applied, etc?  If you instead take a centered, say 12-hour period, do you get the same 1.4cm precision?

Figure 3: Suggestion: In the caption indicate that the subplots are arranged by OGRE location from west to east. Adding a bold title or something similar to the 879* stations to indicate they are the stations along kinematic surveys would also be helpful for the reader.

Figure 4 caption: typo: “for clarify” -> “to clarify?”

Can you put a point that matches the line color to mark these monthly observations? I agree the line is good for visual continuity but the points 

The up and downward pointing triangles are very hard to see. It appears they are centered on the line? Maybe offsetting these triangles either above or below all stations would make them more visible? Also maybe change the colors of some symbols specifically the x’s marking Spot 3 and 4 which are difficult to see, particular the grey x or where there are overlaps. If the station colors are changed to a more muted color palette the symbols may be more easily seen? Regarding the “detected blowing snow” in particular, if present, blowing snow should be occurring across the entire study area and not necessarily concentrated on a few stations (due to high windspeed and abundance of snow). The presence of blowing snow could therefore be indicated at the top or bottom of the graph at each time period (by a symbol or shading vertically at that tilmestep) which would reduce some visual clutter. 

L204: do you mean “Moreover”?

L218: “would” between “but” and “also”

General comments

This is a great study. We need regular altimetry validation studies that use observations from kinematic GNSS, and this study which uses a unique dataset that samples surface height change in winter and in summer fills an observational gap.  I think the manuscript should be accepted with minor revisions. I also have mostly general comments that the authors can take or leave.



My main criticism is that I think the manuscript could go into more detail about the surface elevation trend that you observe at Summit. Do we see an increase or decrease in mean surface height over the 2-year period you have observations? The density of your observations in space and time provide a unique constraint that I think could be used to describe in more detail the subtle surface elevation change signals that are propagating into the interior from more well documented change from the margins. Could you add in the conclusion or the discussion a short subsection that describes these elevation changes and speaks to the significance of this observation in the context of Greenland/Summit accumulation derived from reanalysis and your observations? Consider adding a section in the discussion:



Interior surface‑height change detected by the OGRE network

Beyond bias characterization, the OGRE time series document a net rise in the interior surface of the Greenland ice sheet during 2022-2024..”



It might be outside the scope of this study, but in the future, could you consider using dual frequency GNSS from ground-based radar surveys and UNAVCO kinematic GNSS data to increase the number of GNSS/IS2 crossover points? Much of this data exists and is fully processed on UNVACO/CReSIS servers already, and though most of this data was collected without monitoring sinking from sleds, the sled design/geometry is well constrained and photos from different seasons could be used  to calculate the sinking term and augment the year-round surveying described here for summer months over different regions of Antarctica and Greenland for environments where the surface conditions are more rough near the coast.



One other question I have is connected to the methodology and processing of the kinematic GNSS data and -IR data. I think with a base station at summit and the network in Greenland maintained by UNAVCO, it should be possible to process the kinematic GNSS using TRACK relative to base station solutions using software like GAMIT/GLOBK (or public solutions from repositories that host GNET and summit data). In the case of the OGRE, processing with GAMIT/GLOBK as part of a larger solution for Greenland may improve the relative surface height estimates and could be worth considering in the future. If you need help setting this up for future studies, we can connect after the review period (I don’t think it affects any of your main conclusions here).



The only other delicate suggestion I have is to perhaps make less strong claims about the originality of the autonomy of these systems. For instance, someone likely from unavco or pascal is raising these sensors to make it possible to do GNSS-IR over multiple seasons. This is a lot of work, and it’s been done for quite a while at Summit, but also more remote sites. For instance, take the second paragraph of the conclusion:



“We also present an autonomous method of retrieving ground-based surface elevation estimates using GNSS interferometric reflectometry with a standard GNSS receiver, mounted on a mast in the snow.”



This language makes it seem as though this is the first use of GNSS-IR for monitoring surface height change of ice sheets when most of the methods you’ve described are well established (and I think still require people to service the instruments?). I think these sentences could be modified to emphasize the novel application (surface tracking/altimetry validation) using an established method. I don’t think This was intentional, and my suggestion is just to make this more clear.

Below are minor suggested changes for style and content:

Figure 1: In panel B, it appears most of the reflections are coming from within this azimuth angle of 5 degrees, but that this zone doesn’t overlap in this case with the icesat2 passes. Could similar figures be made for all sites to show how where the measurements you’re making are relative to the the icesat-2 tracks.

LN: 7-11 consider removing autonomy, and defining GNSS, GNSS-IR. Also choos GNSS or GPS (as I think you probably use solutions from all the satellites not just GPS?)

“while reliable, these surveys are resource-intensive. We introduce an alternative, novel validation method using Global Navigation Satellite System (GNSS) Interferometric Reflectometry…”

L12-16 Consider quoting the bias and standard deviation directly. Revealing mean bias of *** +/- ** cm relative to ICESat-2.



L23-31: “By contributing a complementary geographic setting to antarctica..”



This is a long and awkward sentence to me -consider revising, and reframing around Summit as a legacy validation site and link to icebridge and icesat-1.



LN 112-118 (Section 3.1): “The median standard deviation of each pseudo-static point..



This was a sentence that I felt could be shortened and combined with the second sentence:

The median standard deviation of the pseudo-static points was 0.8 cm (n=****), consistent with prior Summit estimates (0.9 -1.8 cm).

LN 158- 166 (Section 3.4): “We follow the same method described above to compare ICESAT-2…”

Consider chaning for clarity to: For ATL06 – ORGE, we apply the same filters (quality.= 0 ; …, but we use a 60 m search radius (beam-pair spacing 45 m) and a linear cross-track interpolation between Spots 3-4 at the OGRE latitude.

LN 233- 239 (Section 5): Consider changing “When we segregate…” to “Separating overpasses flagged for blowing snow or clouds does not affect bias or precision.”



Appendix A: It looks like there was an idea that was not finished or completed. What was the example the authors intended to include here (e.g. …).



Minor comment (not necessarily for a single section):

Can you include a summary table of the parameters you used in the GNSSrefl code. It would also be great to include figures of the Fresnel zone for each receiver as this shows really explicitly the area that you sample from. Height solutions can be sensitive to fesnel zones and the threshold azimuth angle, and recording all this information in a table could help users who want to replicate this kind of study quickly (a lot of this is already well documented in the code). Also include information about which frequencies (likely both?) and constellations were used in the reflection solution.

Copy edits:

LN 29: bi-monthly -> bimonthly, and consider rewriting for clarity



LN 44: Consider citing Larson & Nievinski (2013), Seigfried et al., (2017), Hoffman et al., (2025), Trine et al., (2024), which have used -IR to measure accumulation. GNSS-IR is a powerful measurement technology that is still underused in the glaciological community. Citing these other studies can bring awareness to this method and how it can be used in validation studies of surface height change and to understand near surface accumulation and firn densification.



LN 190: delete temporal

Figure 4 caption: Clarify -> clarity

Section 5: snow pack -> snowpack.

Throughout: Use ~ throughout to approximate value.

Throughout: Overflight -> overpass .. I’m not sure what the community standard is here. Flight seemed odd to me, but I could be wrong.

Throughout: Sub daily -> subdaily

Throughout: Include space before units.

Throughout: 1-\sigma -> 1\sigma

Throughout: Consider abbreviating y^-1 to yr throughout and being consistent with abbreviation of s y, and day (d). I defer to the editor on this. I’m not sure what the best practices are for the cryosphere.