Machine Learning Classification of the Peak Levels of Spring Ice Run at the Arctic Rivers

Лисенков Иван Александрович; Каинова Светлана Андреевна; Луппов Евгений Олегович; Кузнецов Владислав Александрович

doi:10.2205/2026es001147

ВАК 1.6 УДК 004.852 ГРНТИ 37.01 ОКСО 05.00.00 ББК 26 ТБК 63 BISAC SCI УДК 556 УДК 55 УДК 550.34 УДК 550.383 ГРНТИ 37.15 ГРНТИ 37.25 ГРНТИ 37.31 ГРНТИ 38.01 ГРНТИ 36.00 ГРНТИ 37.00 ГРНТИ 38.00 ГРНТИ 39.00 ГРНТИ 52.00

Ещё 20 Свернуть

MACHINE LEARNING CLASSIFICATION OF THE PEAK LEVELS OF SPRING ICE RUN AT THE ARCTIC RIVERS Machine Learning Classification of the Peak Levels of Spring Ice Run at the Arctic Rivers

АА Аа

Опубликовано в Russian Journal of Earth Sciences · Том 26, Номер 3 · Номер статьи: ES3016 · Рубрика: ОРИГИНАЛЬНЫЕ СТАТЬИ

DOI: https://doi.org/10.2205/2026es001147 · EDN: ESQYLD

Получено: 02.06.2026 Одобрено: 02.07.2026 Опубликовано: 31.08.2026 Язык публикаций: RUS

Авторы

ЛИ

Лисенков Иван Александрович

ORCID

КА

Каинова Светлана Андреевна

ORCID

ЛУ

Луппов Евгений Олегович

ORCID

КУ

Кузнецов Владислав Александрович

ORCID

ЛИ

Лисенков Иван Александрович

Геофизический центр Российской Академии Наук Москва

Национальный исследовательский ядерный университет МИФИ Россия

ORCID

КА

Каинова Светлана Андреевна

Геофизический центр Российской Академии Наук Москва

ORCID

ЛУ

Луппов Евгений Олегович

Геофизический центр Российской Академии Наук Москва

Российский университет транспорта Россия

ORCID

КУ

Кузнецов Владислав Александрович

Национальный исследовательский ядерный университет МИФИ Россия

ORCID

АННОТАЦИЯ

Floods during spring ice run are among the most hazardous hydrological phenomena at the rivers of the northern part of the European territory of Russia (ETR). This paper presents an original approach to classifying the peak levels of spring ice run against the critical water level thresholds (markers of adverse and hazardous phenomena) during spring ice run in the lower reaches of the Onega, the Northern Dvina, the Mezen, and the Pechora rivers. This approach is based on the consolidation of heterogeneous data in spatiotemporal neighborhoods and the application of machine learning methods. A comprehensive dataset, which includes long-term hydrological and meteorological observation series for 1961–2023, channel and catchment morphometric characteristics, and ice regime indicators, aggregated within a circular neighborhood with a radius of 25–250 km around 29 hydrometric stations, has been developed. On the basis of the analysis of hydrological data of 7 locations in the lower reaches of the studied rivers, the effectiveness of classical models of machine learning for binary classification of hydrological situations is demonstrated (“no adverse or hazardous phenomenon” / “adverse or hazardous phenomenon”). Five machine learning model training strategies have been tested (general training, spatial generalization across basins and stations, and isolated training). Isolated training was found to provide the best classification quality for three of the four studied basins (Onega, Northern Dvina, Mezen; F1-macro 0.684–1.000), while model transferring between river basins has demonstrated lower quality (F1-macro 0.255–0.608) due to the high specificity of local hydrological conditions. Spatial generalization at the station level within a single basin (F1-macro 0.684–0.774) is recommended when observations are limited. Feature importance analysis revealed the key role of temperature, precipitation, and hydrological predictors. The developed methods can be useful in planning emergency prevention measures.

КЛЮЧЕВЫЕ СЛОВА

systems analysis, geospatial data, multiparametric analysis, data consolidation, ice run, ice jams, hydrological forecasting, risk management

Финансирование

This work was conducted in the framework of budgetary funding of the Russian University of Transport (Project No. 103-00001-26-00, dated January 15, 2026). This work employed facilities and data provided by the Shared Research Facility “Analytical Geomagnetic Data Center” of the Geophysical Center of RAS (http://ckp.gcras.ru/).

Текст

Текст (PDF)

Читать Скачать

1. Agafonova S. A., Frolova N. L., Vasilenko A. N., et al. Ice regime and dangerous hydrological phenomena on rivers of the Arctic zone of European Russia // Lomonosov Geography Journal. — 2016. — No. 6. — P. 41–49. — (In Russian).

2. Agafonova S. A. and Vasilenko A. N. Hazardous Ice Phenomena In Rivers Of The Russian Arctic Zone Under Current Climate Conditions And The Safety Of Water Use // Geography, Environment, Sustainability. — 2020. — Vol. 13, no. 2. — P. 43–51. — https://doi.org/10.24057/2071-9388-2020-12

3. Blaskey D., Racine I., Harlan M., et al. Using Remote Sensing, Statistical, and Machine Learning Techniques to Assess Alaskan River Ice Phenology and Thickness. — ESS Open Archive [Preprint], 2024. — https://doi.org/10.22541/essoar.172954196.61576186/v1

4. Breiman L. Random Forests // Machine Learning. — 2001. — Vol. 45, no. 1. — P. 5–32. — https://doi.org/10.1023/a:1010933404324

5. Buzin V. A. Ice Frazil and Ice Jams on the Rivers of Russia. — St. Petersburg : Es Pe Kha Publ., 2015. — 240 p. — (In Russian).

6. Elreedy D. and Atiya A. F. A Comprehensive Analysis of Synthetic Minority Oversampling Technique (SMOTE) for handling class imbalance // Information Sciences. — 2019. — Vol. 505. — P. 32–64. — https://doi.org/10.1016/j.ins.2019.07.070

7. Gvishiani A. D., Dzeboev B. A. and Agayan S. M. FCaZm intelligent recognition system for locating areas prone to strong earthquakes in the Andean and Caucasian mountain belts // Izvestiya, Physics of the Solid Earth. — 2016. — Vol. 52, no. 4. — P. 461–491. — https://doi.org/10.1134/s1069351316040017

8. Gvishiani A. D., Dzeboev B. A., Agayan S. M., et al. Fuzzy Sets of High Seismicity Intersections of Morphostructural Lineaments in the Caucasus and in the Altai-Sayan-Baikal Region // Journal of Volcanology and Seismology. — 2021. — Vol. 15, no. 2. — P. 73–79. — https://doi.org/10.1134/s0742046321020032

9. Hawker L., Uhe P., Paulo L., et al. A 30 m global map of elevation with forests and buildings removed // Environmental Research Letters. — 2022. — Vol. 17, no. 2. — P. 024016. — https://doi.org/10.1088/1748-9326/ac4d4f

10. Iglin S. M., Moreydo V. M. and Golovnin K. I. Forecasting rare hydrological events by machine learning methods: case study of ice jams on the Pechora river // Lomonosov Geography Journal. — 2025. — Vol. 80, no. 1. — P. 87–97. — https://doi.org/10.55959/msu0579-9414.5.80.1.6 — (In Russian).

11. Kozlov D. V., Buzin V. A., Frolova N. L., et al. Hazardous Ice Phenomena on Rivers and Reservoirs of Russia / ed. by D. V. Kozlov. — Moscow : Russian State Agrarian University – Moscow Timiryazev Agricultural Academy, 2015. — 348 p. — (In Russian).

12. Lehner B. and Grill G. Global river hydrography and network routing: baseline data and new approaches to study the world’s large river systems // Hydrological Processes. — 2013. — Vol. 27, no. 15. — P. 2171–2186. — https://doi.org/10.1002/hyp.9740

13. Limber R., Hoffman F. M., Schwenk J., et al. Long Short-Term Memory Model to Forecast River Ice Breakup Throughout Alaska USA // Water Resources Research. — 2025. — Vol. 61, no. 9. — e2025WR040635. — https://doi.org/10.1029/2025wr040635

14. Lisenkov I. A., Soloviev A. A., Kuznetsov V. A., et al. Generalized dataset of geological and geophysical information on the eastern sector of the Russian Arctic for machine learning-based analysis // Russian Geology and Geophysics. — 2025. — Vol. 66, no. 2. — P. 210–223. — https://doi.org/10.2113/rgg20244747

15. Madaeni F., Lhissou R., Chokmani K., et al. Ice jam formation, breakup and prediction methods based on hydroclimatic data using artificial intelligence: A review // Cold Regions Science and Technology. — 2020. — Vol. 174. — P. 103032. — https://doi.org/10.1016/j.coldregions.2020.103032

16. Malygin I. V. and Aleshin I. M. Forecasting Ice Jams on the Lena River Using Machine Learning Methods // Izvestiya, Atmospheric and Oceanic Physics. — 2022. — Vol. 58, no. 10. — P. 1218–1225. — https://doi.org/10.1134/s0001433822100061

17. Regularities of hidrological processes / ed. by N. I. Alekseevsky. — Moscow : GEOS, 2012. — 733 p. — (In Russian).

18. Semenova N. K., Sazonov A. A. and Krylenko I. N. Predicting the possibility of ice jams formation using machine learning methods // Proceedings of international scientific conference in memory of outstanding Russian scientist Yury Vinogradov. IV Vinogradov Conference Hydrology: From Learning to Worldview. — St. Petersburg : Izdatelstvo VVM, 2020. — P. 358–361. — (In Russian).

19. Shulyakovsky L. G. On ice jams and jam water levels during river breakup // Meteorologiya i Gidrologiya. — 1951. — No. 7. — P. 45–49. — (In Russian).

20. Simonov Yu. A. Systems for preparing and issuing hydrological forecasts at hydrometcenter of Russia // Meteorologiya i Gidrologiya. — 2023. — No. 12. — P. 27–36. — https://doi.org/10.52002/0130-2906-2023-12-27-36 — (In Russian).

21. Soloviev A. A. and Lisenkov I. A. Review and prospects of applying modern approaches to comprehensive geodata analysis for predicting the spatial distribution of geological and geophysical parameters // Geophysical Research. — 2024. — Vol. 25, no. 2. — P. 20–45. — https://doi.org/10.21455/gr2024.2-2 — (In Russian).

22. Soloviev A. A., Novikova O. V., Gorshkov A. I., et al. Recognition of potential sources of strong earthquakes in the Caucasus region using GIS technologies // Doklady Earth Sciences. — 2013. — Vol. 450, no. 2. — P. 658–660. — https://doi.org/10.1134/s1028334x13060159

23. Sumachev A. E., Banshchikova L. S. and Griga S. A. Using the Methods of Neural Network Learning for Peak Water Level Prediction: A Case Study for the Rivers in the Dvina-Pechora Basin // Russian Meteorology and Hydrology. — 2024. — Vol. 49, no. 4. — P. 354–362. — https://doi.org/10.3103/s1068373924040095

24. Varentsova N. A., Kireeva M. B., Kharlamov M. A., et al. Spring river runoff in the European part of Russia: main factors and their estimation. I. Review of previous studies // Hydrometeorological Research and Forecasting. — 2022. — Vol. 2. — P. 92–116. — https://doi.org/10.37162/2618-9631-2022-2-92-116 — (In Russian).

25. White K. D. Review of prediction methods for breakup ice jams // Canadian Journal of Civil Engineering. — 2003. — Vol. 30, no. 1. — P. 89–100. — https://doi.org/10.1139/l02-047

Контент доступен под лицензией Creative Commons «С указанием авторства версии 4.0 Международная»

PDF JATS XML

Главный редактор

Гвишиани Алексей Джерменович

rjes@wdcb.ru

Заместитель главного редактора

Дзебоев Борис Аркадьевич

rjes@wdcb.ru

Ведущий редактор

Кедров Эрнест Олегович

ernest.kedrov@ya.ru

Ответственный секретарь

Карева Наталья Александровна

rjes@wdcb.ru

296

загрузок

Контент доступен под лицензией Creative Commons «С указанием авторства версии 4.0 Международная» Creative Commons

ГОСТ

Лисенков И. А., Каинова С. А., Луппов Е. О., Кузнецов В. А. Machine Learning Classification of the Peak Levels of Spring Ice Run at the Arctic Rivers // Russian Journal of Earth Sciences. 2026. №. 3. ES3016. DOI: https://doi.org/10.2205/2026es001147. EDN: ESQYLD. (дата обращения: 30.09.2026).

MLA

Лисенков Иван Александрович, Каинова С. А., Луппов Е. О., Кузнецов В. А.. "Machine Learning Classification of the Peak Levels of Spring Ice Run at the Arctic Rivers." Russian Journal of Earth Sciences 26.3 (2026): ES3016. DOI: https://doi.org/10.2205/2026es001147. EDN: ESQYLD.

APA

Лисенков И. А., Каинова С. А., Луппов Е. О., Кузнецов В. А. (2026). Machine Learning Classification of the Peak Levels of Spring Ice Run at the Arctic Rivers. Russian Journal of Earth Sciences, 26(3), ES3016. https://doi.org/10.2205/2026es001147 EDN: ESQYLD.

Chicago

Лисенков Иван Александрович, Каинова С. А., Луппов Е. О., Кузнецов В. А.. "Machine Learning Classification of the Peak Levels of Spring Ice Run at the Arctic Rivers." Russian Journal of Earth Sciences 26, no. 3 (2026): ES3016. https://doi.org/10.2205/2026es001147. EDN: ESQYLD.

WIKI

{{cite journal | author = Лисенков И. А., Каинова С. А., Луппов Е. О., Кузнецов В. А. | title = Machine Learning Classification of the Peak Levels of Spring Ice Run at the Arctic Rivers | journal = Russian Journal of Earth Sciences | year = 2026 | volume = 26 | issue = 3 | pages = ES3016 | doi = 10.2205/2026es001147 }} EDN: ESQYLD

Скопировать

Экспорт: BibTeX EndNote RefMan RusMARC

MACHINE LEARNING CLASSIFICATION OF THE PEAK LEVELS OF SPRING ICE RUN AT THE ARCTIC RIVERS Machine Learning Classification of the Peak Levels of Spring Ice Run at the Arctic Rivers

Подтверждение