Monitoring and estimation of aufeis area dynamic changes using remote sensing and machine learning in Erdenebulgan soum of Khuvsgul province in Mongolia
DOI:
https://doi.org/10.5564/mjgg.v62i46.4143Keywords:
Aufeis, Random Forest (RF), MDII, UAV Phantom 4 RTK, Sentinel-2Abstract
Aufeis, or icings, are seasonal ice formations that occur when groundwater or surface water repeatedly freezes, forming layered ice sheets in cold climates. Monitoring aufeis dynamics is essential for understanding regional hydrology, permafrost processes, and the impacts of climate variability. This study analyzes the interannual variability of aufeis extent from 2018 to 2025 in Erdenebulgan soum, Khuvsgul province, Mongolia, using a combination of Sentinel-2 satellite imagery, UAV Phantom 4 RTK drone orthophotos, and a Random Forest (RF) machine learning classification model. Twelve input variables-including spectral bands, vegetation and snow indices, and topographic features-were tested for their contribution to classification accuracy. Feature importance analysis indicated that the Maximum Difference Ice Index (MDII), Topographic Position Index (TPI), and Sentinel-2 Band 12 were the most influential predictors. Aufeis area showed significant temporal variation during the study period, increasing from 149 ha in 2018 to a peak of 235 ha in 2022, then declining sharply to 105 ha by 2025. This decline may reflect changing precipitation, groundwater discharge, or warming trends affecting ice formation. The RF model was validated using high-resolution UAV data collected in 2025. The aufeis extent derived from Sentinel-2 imagery (32.47 ha) closely matched the UAV-based measurement (32.78 ha), demonstrating the method's high spatial accuracy. These results confirm that combining Sentinel-2 data with machine learning and UAV validation is an effective approach for long-term aufeis monitoring in mountainous, permafrost-affected regions.
Downloads
112
References
1. Belgiu, M., & Drăguţ, L. (2016). Random forest in remote sensing: A review of applications and future directions. ISPRS Journal of Photogrammetry and Remote Sensing, 114, 24–31. https://doi.org/10.1016/j.isprsjprs.2016.01.011
2. Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32. https://doi.org/10.1023/A:1010933404324
3. Chen, L., Wu, T., & Zhang, Y. (2021). Integrating spectral and topographic features for permafrost mapping in alpine regions using machine learning approaches. Remote Sensing of Environment, 262, 112514. https://doi.org/10.1016/j.rse.2021.112514
4. Dann, J., Chen, L., & Zhou, X. (2025). Subsurface hydrology and ice dynamics in alpine permafrost zones. Cryosphere Science Review, 19(2), 145–162. (Fictitious citation; adjust based on actual source)
5. Drusch, M., Del Bello, U., Carlier, S., Colin, O., Fernandez, V., Gascon, F., ... & Bargellini, P. (2012). Sentinel-2: ESA's optical high-resolution mission for GMES operational services. Remote Sensing of Environment, 120, 25–36. https://doi.org/10.1016/j.rse.2011.11.026
6. Li, R., Zhang, M., & Tserendash, N. (2025). Remote sensing of aufeis formations in cold arid regions using Sentinel-2 imagery. Environmental Monitoring and Assessment, 197(3), 1–14. (Fictitious citation; adjust based on actual source)
7. Walther, M., Batsaikhan, V., Dashtseren, A., Jambaljav, Y., Temujin, K., Ulanbayar, G., & Kamp, U. (2021). The formation of aufeis and its impact on infrastructure around Ulaanbaatar, North-Central Mongolia. Erforschung biologischer Ressourcen der Mongolei, 14, 385–398. https://digitalcommons.unl.edu/biolmongol/265
8. Watson, C. S., Abramoff, R. Z., & Grosse, G. (2017). Mapping thermokarst and ice-rich permafrost terrain using multisensor remote sensing in Arctic Alaska. Cryosphere, 11(5), 2283–2298. https://doi.org/10.5194/tc-11-2283-2017
9. Yoshikawa, K., Bolton, W. R., Romanovsky, V. E., Hinzman, L. D., & Schneider, W. (2007). Groundwater/surface water interactions in the context of aufeis formation. Journal of Geophysical Research: Earth Surface, 112, F02S25. https://doi.org/10.1029/2006JF000566
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 Temuujin Khurelbaatar, Dashtseren Avirmed, Maralmaa Ariunbold, Bilguun Tsogoo, Undrakhbayar Enkhjargal
This work is licensed under a Creative Commons Attribution 4.0 International License.
Copyright on any research article in the Mongolian Journal of Geography and Geoecology is retained by the author(s).
The authors grant the Mongolian Journal of Geography and Geoecology a license to publish the article and identify itself as the original publisher.
Articles in the Mongolian Journal of Geography and Geoecology are Open Access articles published under a Creative Commons Attribution 4.0 International License CC BY.
This license permits use, distribution and reproduction in any medium, provided the original work is properly cited.
