UDC 630
UDC 004.8
The article analyzes data on wildfires in the Krasnoyarsk Territory for the period 2011-2023, taking into account their geographical distribution. The article examines: (1) exploratory analysis of fire data, including anthropogenic impact on fire activity and changes in fires over the years; (2) building a machine learning model that predicts whether the heat point is critical, i.e., it is located closer than 5 km from a populated area or the fire area is more than 200 hectares. The novelty of the work lies in the effective use of the R programming language, endowed with advanced capabilities for data analysis, modeling by machine learning and visualization methods. The use of data processing methods discussed in the article can improve the accuracy of detecting and predicting changes in fire activity, subsequent monitoring and large-scale firefighting management. The approaches discussed in the article can be applied to other regions of the Russian Federation
wildfires, machine learning, modeling
1. EMISS. Oficial'nye statisticheskie pokazateli [Elektronnyy resurs]. - Rezhim dostupa: https://fedstat.ru.
2. Ponomarev E.I., Haruk V.I., Yakimov N.D. Rezul'taty i perspektivy sputnikovogo monitoringa prirodnyh pozharov // Sibirskiy lesnoy zhurnal. 2017. No 5. S. 25-36. EDN: https://elibrary.ru/ZRULTR DOI: https://doi.org/10.15372/SJFS20170503
3. Volokitina A.V., Sofronova T.M., Korec M.A. Regional'nye shkaly ocenki pozharnoy opasnosti v lesu: usovershenstvovannaya metodika sostavleniya // Sibirskiy lesnoy zhurnal. 2017. No 2. S. 52-61. EDN: https://elibrary.ru/YODBDT DOI: https://doi.org/10.15372/SJFS20170206
4. ISDM-Rosleshoz. Informacionnaya sistema distancionnogo monitoringa Federal'nogo agentstva lesnogo hozyaystva [Elektronnyy resurs]. - Rezhim dostupa: https://nffc.aviales.ru/main_pages/index.shtml.
5. The R Project for Statistical Computing [Elektronnyy resurs]. - Rezhim dostupa: https://www.r-project.org.
6. Robin Lovelace, Jakub Nowosad, Jannes Muenchow. Geocomputation with R. Chapman Hall/CRC, 2019. DOI: https://doi.org/10.1201/9780203730058
7. Samsonov T.E. Vizualizaciya i analiz geograficheskih dannyh na yazyke R. M.: Geograficheskiy fakul'tet MGU, 2022.
8. Sergio J. Rey, Dani Arribas-Bel, Levi John Wolf. Geographic Data Science with Python. Chapman Hall/CRC, 2023.
9. Tidyverse. The R packages for data science [Elektronnyy resurs]. - Rezhim dostupa: https://www.tidyverse.org.
10. Tidymodels: the collection of packages for modeling and machine learning using tidyverse principles [Elektronnyy resurs]. - Rezhim dostupa: https://www.tidymodels.org.
11. Babenyshev S.V. i dr. Analiz pozharnoy i tehnosfernoy bezopasnosti gorodskoy sredy g. Krasnoyarska s primeneniem geoprostranstvennyh instrumentov // Sibirskiy pozharno-spasatel'nyy vestnik. 2021. No 4. S. 128-136. EDN: https://elibrary.ru/JTQQLP DOI: https://doi.org/10.34987/vestnik.sibpsa.2021.21.69.001
12. Babenyshev S.V. i dr. Primenenie geoinformacionnyh instrumentov dlya raboty s bol'shimi dannymi pri analize prostranstvennogo raspredeleniya pozharov // Sibirskiy pozharno-spasatel'nyy vestnik. 2021. No 2. S. 70-77. EDN: https://elibrary.ru/YOJATK DOI: https://doi.org/10.34987/vestnik.sibpsa.2021.47.69.013
13. Wickham H. i dr. R for Data Science, 2nd edition. 2023.
14. Kuhn M., Silge J. Tidy Modeling with R: A Framework for Modeling in the Tidyverse. O'Reilly Media, 2022.
15. Ob utverzhdenii Pravil tusheniya lesnyh pozharov: prikaz Ministerstva prirodnyh resursov i ekologii RF ot 1 aprelya 2022 g. No 244 [Elektronnyy resurs]. 2022. - Rezhim dostupa: https://www.garant.ru/products/ipo/prime/doc/405045151.
16. Spatialsample: a library to provide building blocks for creating and analyzing resamples of a spatial data [Elektronnyy resurs]. - Rezhim dostupa: https://spatialsample.tidymodels.org.
17. Parsnip: Search parsnip models [Elektronnyy resurs]. - Rezhim dostupa: https://www.tidymodels.org/find/parsnip.
18. Shitikov V.K., Mastickiy S.E. Klassifikaciya, regressiya i drugie algoritmy Data Mining s ispol'zovaniem R. 2017. EDN: https://elibrary.ru/ZVTCRF
