DEVELOPMENT OF A SYSTEM FOR AUTOMATED FORECASTING OF AVALANCHE DANGER, USING MODERN ICT AND THE APPARATUS OF NEUROCOGNITIVE ARCHITECTURES

Authors

  • Rinat I. Pupcev Federal State Budget-Financed Educational Institution of Higher Education The Bonch-Bruevich Saint Petersburg State University of Telecommunications, St. Petersburg, Russia

DOI:

https://doi.org/10.34753/HS.2022.4.3.244
+ Keywords

machine learning, avalanche forecasting, snow avalanches, ro service architecture, data analysis, data collection

+ Abstract

The development of infocommunication technologies opens up great opportunities for the creation of avalanche hazard forecasting systems, the rapid development of containerization and virtualization technologies make it possible to fully use the concept of microservice architecture. The development of the machine learning apparatus opens up wide opportunities for its application to solve the forecasting problem. The use of proven methods can give good results and increase the accuracy of avalanche hazard forecasts. This article discusses: the problem of using modern information and communication technologies and machine learning in the field of avalanche forecasting, data sources for the functioning of an automated avalanche hazard forecasting system. The most important data for use in the operation of such a system are highlighted. There is also a list of requirements that such a system must meet. A prototype of a system for predicting the background level of avalanche danger for the mountain region of Khibiny has been developed. The prototype works using machine learning algorithms. A set of predicted meteorological parameters is used for the operation of the models, the forecast is made for the next day and updated every hour. In addition, an analytical model was launched to predict avalanches from wind slab. At the end of the work, examples of displaying data obtained as a result of machine learning models are given, an example of a message sending information by a Telegram bot is given. Separately, it is necessary to highlight the possibility of integrating the system with existing data collection and analysis systems, the introduction of already proven systems will expand their functionality and provide new opportunities for avalanche forecasters.

+ Author Biographies

Rinat I. Pupcev, Federal State Budget-Financed Educational Institution of Higher Education The Bonch-Bruevich Saint Petersburg State University of Telecommunications, St. Petersburg, Russia

eLibrary (РИНЦ)
SPIN-
код

ORCID ID

7176-4840

0000-0003-2145-4274

+ References

Akkuratov V.N. Meteorologicheskie usloviya obrazovaniya lavin v Khibinakh [Meteorological conditions for the formation of avalanches in the Khibiny]. Materialy glyatsiologicheskikh issledovanii [Data of Glaciological Studies], 1966, no. 12, pp. 132–138.

Avalanche Bulletin Interpretation Guide. Edition November 2022. WSL Institute for Snow and Avalanche Research SLF (ed.). 2022.

URL: https://www.slf.ch/en/avalanche-bulletin-and-snow-situation/about-the-avalanche-bulletin/Interpretationshilfe_en.pdf.

Bianchi F.M., Grahn J., Eckerstorfer M., Malnes E., Vickers H. Snow avalanche segmentation in SAR images with Fully Convolutional Neural Networks. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2021, vol. 14, pp. 75–82. DOI: 10.1109/JSTARS.2020.3036914.

Blagovechshenskiy V.P., Zhdanov V.V. Opyt otsenki i prognoza lavinnoi opasnosti v Shveĭtsarii [Experience of evaluation and forecast of avalance danger in Switzerland]. Gidrometorologiya i ekologiya [Hydrometeorology and Ecology], 2019, No. 1 (92), pp. 178–190. (In Russian; abstract in English).

Kondrashov I.V. Prognoz lavin i nekotorykh kharakteristik snezhnosti v gorakh Kazakhstana [Forecast of avalanches and some characteristics of snow cover in the mountains of Kazakhstan]. Leningrad, Publ. Gidrometizdat, 1991. 72 p.

(In Russian).

Nowak S., Bartram L., Haegeli P. Designing for Ambiguity: Visual Analytics in Avalanche Forecasting. Proceedings of the International Conference on Visualisation, VIS (Virtual Event, 25–30 October 2020). Los Alamitos, California, Washington, Tokyo: IEEE Computer Society, 2020. DOI: 10.1109/VIS47514.2020.00023.

Solov'ev A.Yu. Geoinformatsionnye metody issledovaniya lavinnoi opasnosti na primere Khibinskogo gornogo massiva. Diss. … kand. geogr. Nauk [Geoinformation methods for studying avalanche danger on the example of the Khibiny mountain range. Ph. D. (Geography) Thesis]. Moscow, Publ. of Moscow State University, 2002. 143 p. (In Russian).

Statham G., Haegeli P., Greene E., Birkeland K., Israelson C., Tremper B., Stethem Ch.,

McMahon B., White B., Kelly J. A conceptual model of avalanche hazard. Natural Hazards, 2018, vol. 90, iss. 2, pp. 663–691. DOI: 10.1007/s11069-017-3070-5.

Techel F., Schweizer J. On using local avalanche danger level estimates for regional forecast verification. Cold Regions Science and Technology, 2017, vol. 177, pp. 55–62.

DOI: 10.1016/j.coldregions.2017.07.012.

Zhdanov V.V. Sovremennaya kontseptsiya razvitiya prognozov lavinnoi opasnosti (na primere prognozov v raione SLS Shymbulak) [Modern concept of development of forecasts of avalanche danger (on the example of forecasts in the area of avalanche station "Shymbulak")]. Gidrometorologiya i ekologiya [Hydrometeorology and Ecology], 2020, No. 2 (97), pp. 83–95.

(In Russian; abstract in English).

Rukovodstvo po snegolavinnym rabotam (vremennoe) [Avalanche Manual (Temporary)]. Leningrad, Publ. Gidrometizdat, 1965. 397 p.

(In Russian).

+ Read article online

Downloads

Download data is not yet available.

Published

2023-04-20

How to Cite

Rinat I. Pupcev. (2023). DEVELOPMENT OF A SYSTEM FOR AUTOMATED FORECASTING OF AVALANCHE DANGER, USING MODERN ICT AND THE APPARATUS OF NEUROCOGNITIVE ARCHITECTURES. Hydrosphere. Hazard Processes and Phenomena, 4(3), 244–254. https://doi.org/10.34753/HS.2022.4.3.244
Loading...