Article name Geoecological Mapping of Buildings in Flood-Prone Areas Shilka and Nerchinsk (Transbaikalia), Using the NDBI Spectral Index and a Neural Network
Authors Kochev D.. ,
Bibliographic description Kochev D. V. Geoecological Mapping of Buildings in Flood-Prone Areas Shilka and Nerchinsk (Transbaikalia), Using the NDBI Spectral Index and a Neural Network // Transbaikal State University Journal. 2024. Vol. 30, no. 1. P. 28–39. DOI: 10.2109/2227-9245-2024-30-1-28-39.
Category Earth and Environmental Sciences
DOI 502/504; 528.88
DOI 10.2109/2227-9245-2024-30-1-28-39
Article type Original article
Annotation At all times, the problem of floods has been relevant due to their destructiveness, frequency and poor predictability. The situation is complicated by the fact that floodplains are used for economic, recreational and other activities, thereby exposing the risk of destruction and degradation of any objects and human waste products during floods. Therefore, mapping and spatiotemporal monitoring of flood-prone areas in order to analyze and predict the likely material and economic damage to facilities of economic activity from the threat of floods is an urgent scientific problem that can be solved by geographic information systems using satellite data processed by spectral indices and convolutional neural networks. The object of the study is the flood-prone territories of Shilka and Nerchinsk (Transbaikal Region). The purpose is to assess the accuracy of economically developed territories decryption in flood-prone areas of Shilka and Nerchinsk in order to determine the effective threshold for separation of the studied facilities. Research objectives are as follows: collection and formation of an archive of Earth remote sensing data of the Landsat-8 OLI program; inference of satellite images on a neural network and their use as a reference image; calculation of NDBI spectral indices based on Landsat-8 data; determination of the areas of buil­­t-up and economically developed territories in the images; assessment of the accuracy and comparative analysis of the results of the development facilities allocation and economic activity; identification of advantages, disadvantages, applicability limits of the methods used to decrypt buildings and objects of economic activity in flood-prone areas. The accuracy and reliability of the results of the decrypting buildings and objects of economic activity by calculating the NDBI spectral index, the inference of a convolutional neural network according to Landsat 8 OLI data in flood-prone areas of Shilka and Nerchinsk have been evaluated. The most effective methods, parameters and thresholds for decrypting buildings and objects of economic activity have been determined in order to monitor the development of flood-prone areas and control the flood-prone situation.
Key words decryption, floods, flood-prone areas, spectral indices, remote sensing of the Earth, Landsat-8 OLI, convolutional neural networks, geoinformation analysis, accuracy assessment, kappa
Article information
References Vakhnina I. L., Obyazov V. A., Zamana L. V. Dynamics of moisture in the steppe zone of South-Eastern Transbaikalia from the beginning of the 19th century based on Scots pine cores. Bulletin of the Moscow University. Episode 5. Geography, no. 2, pp. 28–33, 2018. EDN YUUDHA. (In Rus.) Kalashnikova L. Ya., Stasyuk O. N., Avseenko N. D. Floods in the Transbaikal Territory. Innovative technologies in technology and education: Collected articles of the VIII International Scientific and Practical Conference (Chita, December 8–9, 2016). Chita: Transbaikal State University, 2016. EDN YMCUSN. (In Rus.) Kochev D. V., Kurganovich K. A. Using remote sensing data and artificial intelligence algorithms to analyze flood-prone areas // Evolution of the biosphere and technogenesis: proceedings of the All-Russian Conference with international participation. Chita: IPREK SB RAS, 2022. DOI: 10.57245/978_5_9293_3064_3_2022_2_304. EDN NQHGHU. (In Rus.) Kochev D. V., Shumilova L. V. Application of spectral water indices at the tailings of the Darasun mine according to Earth remote sensing data from the Landsat program. Bulletin of the Transbaikal State University, vol. 29, no. 2, pp. 45–60, 2023. DOI: 10.2109/2227-9245-2023-29-2-45-60. EDN VWZNGB. (In Rus.) Kurbanov E. A., Vorobyov O. N., Gubaev A. V., Lezhnin S. A., Polevshchikova Yu. A., Demisheva E. N. Four decades of forest research using Landsat images. Bulletin of the Volga State Technological University. Series «Forest. Ecology. Nature management», no. 1, pp. 18–32, 2014. EDN RXHHFP. (In Rus.) Kurganovich K. A., Shalikovsky A. V., Bosov M. A., Kochev D. V. Application of artificial intelligence algorithms for monitoring flood-prone areas. Water management of Russia: problems, technologies, management, no. 3, pp. 6–24, 2021. DOI: 10.35567/1999-4508-2021-3-1. EDN NIEEKC. (In Rus.) Faizulloev Sh. A. Assessment of the accuracy of classification of remote sensing data in GIS technology using the example of the hydroelectric power station cascade area on the Vakhsh River. News of the Academy of Sciences of the Republic of Tajikistan. Department of physical, mathematical, chemical, geological and technical sciences, no. 1, pp. 96–103, 2020. EDN MHZPOX. (In Rus.) Shalikovsky A. V. Flood 2018 in the Trans-Baikal Territory. Kulagin readings: equipment and technologies of production processes: XVIII International scientific and practical conference: in 3 parts (Chita, November 28–30, 2018). Chita: Transbaikal State University, 2018. EDN YUASDZ. (In Rus.) Shalikovsky A. V. Floods in the Transbaikal Territory: causes, consequences, forecast possibilities. Water resources and water use: collection of works (Chita, July 20, 2019). Chita: Transbaikal State University, vol. 9, 2019. EDN ZVLRFH. (In Rus.) Shalikovsky A. V. Fundamentals of rational use of flood-prone areas. Abstract of Doctor of Geographical Sciences. Chita, 2004. EDN NHVWPB. (In Rus.) Feizizadeh B., Darabi S., Blaschke T., Lakes T. QADI as a New Method and Alternative to Kappa for Accuracy Assessment of Remote Sensing-Based Image Classification. Sensors, vol. 22, no. 12, pp. 4506, 2022. DOI: 10.3390/s22124506. EDN TBUBTP. (In Eng.) Krizhevsky A., Sutskever I., Hinton G. E. ImageNet classification with deep convolutional neural networks. Web. 08.11.2023.­volu­tional-neuralnetworks.pdf. (In Eng.) Qiu S., Lin Y., Shang R. Making Landsat time series consistent: Evaluating and improving Landsat analysis ready. Remote Sensing, vol. 11, no. 1, pp. 51, 2019. DOI: 10.3390/rs11010051. EDN TFDHVL. (In Eng.) Zha Y., Gao J., Ni S. Use of normalized difference built-up index in automatically mapping urban areas from TM imagery. International Journal of Remote Sensing, vol. 24, no. 3, pp. 583–594, 2023. DOI: 10.1080/01431160304987. (In Eng.) Zhang L., Wu J., Fan Y., Gao H., Shao Y. An Efficient Building Extraction Method from High Spatial Resolution Remote Sensing Images Based on Improved Mask R-CNN. Sensors (Basel), vol. 20, no. 5, 2020. (In Eng.)
Full articleGeoecological Mapping of Buildings in Flood-Prone Areas Shilka and Nerchinsk (Transbaikalia), Using the NDBI Spectral Index and a Neural Network