This paper provides an interesting review of the contrasting behaviour of adjacent surge-type glaciers in the Karakoram, and in particular how a very slowly surging glacier (likely controlled by changes in basal thermal conditions) can exist next to rapidly advancing ones (likely controlled by changes in basal hydrological conditions). These changes are well documented and illustrated through changes in terminus position, surface elevation and ice velocity, both in the main paper and the extensive supplemental material. I haven’t seen many comparisons of surging glaciers which contrast in their behaviour over such a short distance before, so this provides the primary novel contribution of this study. However, the current title doesn’t well describe this: from the title I expected the paper to provide a methodological assessment of the pros and cons of different remote sensing methods for detecting glacier surges (and how some methods might ‘miss’ features due to low spatial or temporal resolution), but the paper doesn’t really do this; the main finding is that most methods actually work pretty well for detecting surges and their associated velocity and elevation changes. So I recommend that the title be changed to something that better describes the actual content of the paper, which if we rearrange the wording from the first sentence of the Conclusions (L833-835) could be something like: ‘Contrasting characteristics and forcing mechanisms of three glacier surges in the central Karakoram’. Rearrangement of a few sentences or sections in the main text to better reflect this focus could help.

My specific comments are below. The biggest one is that the Discussion and Conclusions would benefit from better referencing to previous work.

Detailed comments

L27/28: provide some more information about the ‘full suite of satellite sensors’ and ‘DEMs from different sources’ that you used – e.g., names of sensors, resolution, repeat coverage period. This is one of the most useful things that I would like to know when reading the abstract, more so than some of the other background information currently provided.

L43: fit neatly with what? Clarify

L52: change ‘provides’ to ‘provide’

L72-75: it might be useful to add reference to Hewitt (2005) here, which provides some discussion of the different potential causes of Karakoram surges, particularly in relation to changes in basal thermal regime: Hewitt, K. 2005. The Karakoram Anomaly? Glacier Expansion and the ‘Elevation Effect,’ Karakoram Himalaya. Mountain Research and Development, 25(4), 332-340.

L107: I think that you mean 35.94N, not 5.94N!

L163: seems to be missing ‘with’ before ‘the Landsat 7…’

L173: I don’t really understand what ‘virtual’ refers to here. Either needs more explanation, or stating something like ‘we digitized likely maximum extents’ would seem to be clearer.

L177: What does ‘spatially consistent’ refer to? E.g., orthorectified?

L168-201: For readers who may be unfamiliar with the satellite sensors used, please make it clear as to which are optical and which are SAR. Also provide the resolution for the optical scenes (you currently only mention this for SAR).

L188: please state whether you selected different image types from particular times of the year. E.g., was SAR data preferentially acquired in the winter? Was optical data preferentially collected under summer snow-free conditions?

L207: the SPOT5 DEM is stated as being from January 2010 here, but the date is listed as 31 Oct 2010 in Table 2

L216: Hugonnet et al. (2021) list their datasets as being from 2000-2004 and 2005-2009

L244: state why a two-pixel change is important. Related to the above comment, the resolution of the image sources is also needed to understand what two pixel relates to in real world distance for different image types.

L260: when you mention the offset-tracking here, are you referring to the TSX data? Please clarify

L295: These dates are difficult to decipher. Do you mean ‘4 Nov. and 16 Nov., 2020’? If so, then write out the dates like this.

L310-312: this statement requires better justification for ignoring microwave penetration in the SRTM DEM. Gardelle et al. (2012) state that it averages 3 m in the Karakoram, so this should be compared to the elevation changes caused by the surges.

L363: since you have information concerning the retreat rates of South Chongtar and NN9 prior to their recent surges (e.g., from Fig. S7), then I wonder if it would be useful to plot these on Fig. 4a? Seeing their rate of retreat would be useful to compare to their rate of advance, and help to support the retreat patterns that you describe later in this section.

L557: I’m unclear as to what ‘compensation effects’ is referring to here

L579: probably useful to remind the reader here and earlier in this section as to which sensors are optical (Landsat, Planet, Sentinel 2), and which are SAR (TSX, Sentinel 1).

L596: I’m unclear as to the date of the HMA 2015 DEM as the label at the top of Fig. S12 specifies ‘spring 2015’, but Table 2 indicates Feb.-Aug. 2015. Please clarify.

L661: provide ref(s) to support the statement that these are likely the highest and lowest surge velocities in the Karakoram

L674: clarify as to exactly what characteristics you’re referring to when you say that these glaciers ‘developed nearly all characteristics of a surge’

L697: can you refer to any modelled mass balance data from this region to indicate whether positive mass balance conditions have occurred there recently? That could help to support or refute your statements about potential causes of the observed glacier changes.

L659-740: overall this section could do with better referencing to existing literature. For example, you refer several times to the characteristics of hydrologically controlled vs. thermally controlled surges, but don’t refer to the previous literature which describes this (both for the Karakoram and elsewhere) and the evidence for the different types of flow patterns and lengths of active and quiescent phases. There are many ‘classic’ papers which can be useful here, such as Sevestre and Benn (2015), Murray et al. (2003), Benn et al. (2019; https://doi.org/10.1017/jog.2019.62), and Kamb et al. (1985; https://doi.org/10.1126/science.227.4686.469).

L748: it’s confusing to have the (3) in brackets here; just write out the sentence to describe Sentinel-2 and Planet separately

L755: change ‘relative’ to ‘relatively’

L758: change ‘reliable’ to ‘reliably’

L765 & L810: a major difference between the sensors that you haven’t mentioned is that some are optical and some are SAR, so it would be useful to discuss whether this has any impact on the measured velocities and how suitable different scene pairs are for deriving them. E.g., effect of snow cover, effect of surface melt, SAR layover/foreshortening effects, SAR penetration, optical shadows?

L793: change word order to ‘not yet’

L831-880: similar to my comment above for L659-740, the Conclusions would benefit from better referencing to previous studies so that it’s clear how your findings compare to those of others. At the moment you don’t have a single reference in this section!

L875: would be more accurate to say: ‘…the termini of South Chongtar and NN9 are now colliding”

Tables and Figures

Table 3: Add reference to Fig. S5 in the caption so that it’s clear as to which areas were used for the volume calculations

Fig. 1: I would like to see distance markers along the centrelines, so that it’s easier to understand how the distances shown in other figures (e.g., Figs. 6, 7) relate to physical locations along the glaciers

Fig. 2: It would be useful to split the lines into two groupings: those that rely on DEMs (upper), and those that rely on optical or SAR images (bottom), such as by adding a blank line between them and changing the ‘Sensors’ label on the y-axis to something more descriptive. This is because some of the DEM sources (e.g., ASTER, SPOT) can also provide optical images, but you don’t use images of that type in this study.

Fig. 4: increase the font size for the legend

Figs. 5, 8, 11: the velocities would be easier to see if you cropped out the data to only show it over glaciers, and not over bedrock

Fig. 6: I find it hard to understand the dates of some of these velocity profiles as the colours are pretty similar in the legend (the Landsats are all a pretty similar blue, the more recent S2s all a similar red). Can you use a better colour scale to help separate them, or perhaps used dashed lines for some periods?

Fig. 9: Add a reference to Fig. 1 in the caption to indicate where the cross-profile is located. Also misspelling of ‘Grey numbers…’

Supplemental

L12: Sentine-2 is misspelled

Table S1: please use the format yyyy/mm/dd for the dates so that they’re consistent with Table 2. Also use the same font size throughout.

Fig. S2 and others: state in the caption as to what coordinate system is used for these plots (e.g., UTM 34N?)

General comments

Authors report on the surge behaviour of three glaciers in the same valley discussing differences in timing, extent and progression of the surges. For this purpose, they use a comprehensive dataset comprising different optical and radar satellites to map glacier advance, flow velocities and elevation changes.

The paper is well written with high quality figures and it also gives a good overview on the literature about glacier surges.

There is only one aspect that should be elaborated more carefully: In addition to the analysis of the surge behaviour of the investigated glaciers, authors compare results from different sensors in order to discuss their suitability for similar studies. This intercomparison has to be done in a more quantitative way. Statements like "average differences are insignificant" or "agree very well" are not enough in this context and error measures should be given (see specific comments).

I suggest to accept the manuscript after minor revisions.

Specific comments

line 212 Could you please briefly state what kind of DEM datasets?

line 507 “some surface lowering”: How much?

line 562 Which area was used to calculate the volume loss of Sarpo Laggo Glacier? In Fig. 11 a) and b) there is volume gain visible in the part of the glacier that is not covered in Fig. 11 c and d.

Section 5.6 Uncertainties should be placed before 5.5 Sensor intercomparison. As a result, the differences between the sensors can be quantitatively compared to the determined uncertainties.

line 572 Give error measures and compare with the sensor uncertainties derived in 5.6.

line 597 “agree very well”: Compare it quantitatively with the stable terrain accuracies (Table S3).

line 630 MAD: Table S2 only lists mean and std. Either explain how you derive std from MAD or adjust text and table. Check all numbers cited in 5.6.2 accordingly.

line 667 Explain why “maximum flow velocities in summer” are atypical.

line 740 Can you comment on the influence of crevassing on the volume determination? Are the crevasses resolved by the DEMs?  Does the DEM of the crevassed surface represent a mean elevation or is it more like an envelope curve?

line 770 basically identical: But they are from different seasons and Fig. S12 shows the differences. I suggest to give median and MAD to quantify this statement.

line 858 Was Sentinel-1 used in Section 5.1 or only in the animation? Sentinel-1 images are not listed in Table S1.

Technical corrections

line 27 same region: I'd narrow it down to the same valley, even.

line 43 Is it ICESat-2 ATL03?

line 102 by Hugonnet et al. (2021): Please make it clearer that this is an external dataset.

line 107 35.94°N

line 139 At their historically recorded maximum extent

line 181 Sentinel-1 is not included in table S1.

line 295 Correct date format.

line 299 Which standard processing steps? If these are listed in the following passage, use a colon (:) at the end of this phrase.

line 305 "We also excluded data voids": Seems clear. How could you alternatively include voids?

line 438 Replace "afterwards" with "further down".

line 450 Fig. 7a confirms this observation:

line 451 "(up to 0.2 m d-1)":  Two lines above the maximum is 0.4 m/d.

line 472 Replace "3 km" with "km 3" like in line 474

line 474 What do you mean with "maximum velocities near km 15 changed only slowly over the entire glacier length"? Do you refer to the location km 15 or the entire length?

line 475 Replace "10 km" with "km 10".

lines 590, 592, 597: Specify subfigure in Fig. S6 (see comment on S6 below).

line 643 median elevation differences to the reference DEM “on bedrock”

line 670 Replace “implies” with “imply”.

line 708 ”location of the maximum away from the centre”: Do you mean centre line?

line 764 strictly or strict?

line 799 short-term variation

line 834 in the same valley?

line 836 Also give the duration of the active phase in comparison to North Chongtar.

Table 3, line 1150 numbers

Figure 1 Add map grid

In some figures (e.g. Fig. 6, 7) colours are hard to distinguish. Please check whether you could use colours that are better distinguishable.

Figure 9 State direction of cross profile. West to east?

line 1219 with respect to

Supplemental Material

Table S1: Add Sentinel-1.

Table S3: Why are some lines in italics? I suggest to add NMAD.

Fig. S1: Check dates, resolutions and dataset names as they are inconsistent with Table 2. Subfigure d) Use the same clipping as in a)-c).

Fig. S5, line 60: What are "biased accumulation areas"?

Fig. S6: Consider numbering all panels as subfigures. It makes it easier to refer to a specific panel in the text.

line 66 2 February 2019: Does not coincide with date on map. Shouldn't it be December?

Fig. S7: clockwise? The main paper has the usual order from left to right (e.g. Fig. 5). Please adjust.

Fig. S13: Dates (and order of dates) do not coincide with the ones in the figure for a) to d).