| 27 Mar 2026
High-resolution mapping of glacial lake expansion in Kyrgyzstan (2016–2024) using Sentinel-2 imagery
Abstract. Glacial lakes are sensitive indicators of glacier retreat and are important for regional hydrological and hazard assessments. This study presents a high-resolution (10 m) inventory of glacial lakes in Kyrgyzstan for 2022–2024, derived from Sentinel-2 imagery using a semi-automated workflow complemented by manual refinement. A Random Forest classification model trained on the 2022–2024 inventory was applied to 2016–2017 imagery to reconstruct historical lake distributions and analyze changes over an eight-year period. The 2024 inventory includes 2 592 lakes covering 77.6 km2, mostly small (<0.05 km2) lakes located between 3 500 and 3 800 m elevation. Between 2016 and 2024, the number of lakes increased by 10.5 % and total lake area by 8.7 %, driven primarily by the formation of small, high-altitude proglacial lakes, whereas larger, lower-elevation lakes remained largely stable. Supraglacial lakes exhibited slight area increases and, on average, an upward shift in elevation, whereas proglacial, and glacier-detached lakes showed minimal changes. Regional trends reveal pronounced heterogeneity, with Issyk-Kul, Batken, and Talas emerging as regions of new lake formation. Comparison with global datasets confirms completeness and reliability of our inventories. These results highlight the ongoing influence of glacier retreat on lake formation and expansion in Kyrgyzstan, providing a robust baseline for hazard assessment, water resource management, and future cryospheric monitoring.
This looks like a well-written and interesting study of glacial lake evolution across Kyrgyzstan from 2016-2024. I have only very minor comments on the manuscript.
L6 Abstract. I don’t disagree that lakes are sensitive indicators of glacier change, but glaciers themselves are sensitive indicators of glacier change. You don’t need to look at the lakes necessarily. And also, glacial lakes are not important for hydrological and hazard assessments – they are important considerations or to be considered. This first line needs to say something useful and meaningful and I don’t think it’s quite there yet.
L11-12. Where it starts “mostly small” feels like it should be worked into a separate sentence or after a semi-colon. Extra words needed – something like, “Most of these lakes are relatively small…”
L35-36. The statistics here on the number of glacial lakes in Kyrgyzstan and their susceptibility to GLOFs is based on a paper from 2010 (so the actual date likely earlier than that). So these stats are at least 16 years old, and yet you use the present tense (“are”) when discussing them. Would it not be better for you to make that clearer
Figure 1- I see the urban areas feature in the legend, but can’t see any on the actual map and the colour is too similar anyway to the low elevation shading. Seems to me that red dots were used to mark urban areas anyway.
L95-98 – it would be good to see some citations to back up the claims about changing snow cover, GLOFs, landslides, etc.
L102 – again, would be good to see a citation to substantiate this point about rural livelihoods.
Fig 10 – I didn’t really understand the legend in part b “Area change to 2016 (%)”. Should this be since 2016?