Mapping Siberian Arctic Mountain Permafrost Landscapes by Machine Learning Multi-Sensors Remote Sensing: Example of Adycha River Valley

• Main Authors: Zakharov, Moisei, Gadal, Sébastien, Danilov, Yuri, Kamičaitytė, Jūratė
• Other Authors: Aix Marseille Université (AMU), Étude des Structures, des Processus d’Adaptation et des Changements de l’Espace (ESPACE), Université Nice Sophia Antipolis (1965 - 2019) (UNS)-Avignon Université (AU)-Aix Marseille Université (AMU)-Centre National de la Recherche Scientifique (CNRS)-Université Côte d'Azur (UniCA), North-Eastern Federal University, Kaunas University of Technology (KTU), Polar Urban Centers PUR, INSTICC, SCITEPRESS: Science and Technology Publications, FMSH-RBSF OSAMA (development Of an optimal human Security Model for The Arctic), PEPS CNRS RICOCHET, Cédric Grueau, Robert Laurini, Lemonia Ragia
• Format: Conference Object
• Language: English
• Published: HAL CCSD 2021
• Subjects: Permafrost Landscape; Remote Sensing Modeling; Landscape mapping; Digital Terrain Model; Permafrost map dynamics; Machine learning; Data fusion; Geographic databases; Local knowledge; ASTER GDEM; Landsat 8 OLI / TIRS; Sentinel 2; Multi-sensors; Arctic Mountains; Yakutia; Russia; [SHS.STAT]Humanities and Social Sciences/Methods and statistics; [SDE.ES]Environmental Sciences/Environment and Society; [SHS.GEO]Humanities and Social Sciences/Geography; [INFO.INFO-MO]Computer Science [cs]/Modeling and Simulation; [INFO.INFO-TI]Computer Science [cs]/Image Processing [eess.IV]; Adycha; Arctic; Verkhoyansk; permafrost; Tundra; Siberia
• Online Access: https://hal.science/hal-03207301; https://hal.science/hal-03207301/document; https://hal.science/hal-03207301/file/GISTAM2021_22_Final_.pdf

International audience The landscape taxonomy has a complex structure and hierarchical classification with indicators of their recognition, which is based on a variety of heterogeneous geographic territorial and expert knowledge. This inevitably leads to difficulties in the interpretation of remote sensing data and image analysis in landscape research in the field of classification and mapping. This article examines an approach to the analysis of intra-season Landsat 8 OLI images and modeling of ASTER GDEM data for mapping of mountain permafrost landscapes of Northern Siberia at the scale of 1: 500,000 as well as its methods of classification and geographical recognition. This approach suggests implementing the recognition of terrain types and vegetation types of landscape types. The 8 types of landscape have been identified by using the classification of the relief applying Jenness's algorithm and the assessment of the geomorphological parameters of the valley. The 6 vegetation types have been identified in mountain tundra, mountain woodlands, and valley complexes of the Adycha river valley in the Verkhoyansk mountain range. The results of mapping and the proposed method for the interpretation of remote sensing data used at regional and local levels of studying the characteristics of the permafrost distribution. The work contributes to the understanding of the landscape organization of remote mountainous permafrost areas and to the improvement of methods for mapping the permafrost landscapes for territorial development and rational environmental management.