Scholarly record
IMPORTANCE OF GEOMORPHOMETRIC AND GEOLOGIC VARIABLES FOR AUTOMATED DELINEATION OF GEOMORPHOLOGICAL UNITS (ON THE EXAMPLE OF YAMAL PENINSULA)
Abstract
This study assesses the efficiency of various geomorphometric terrain characteristics and a number of categorical geological parameters in training statistical models to create geomorphological maps. The study area is the central sector of the western coast of Yamal Peninsula (S = 10 400 km2). The terrain is exceptionally flat and morphologically similar, but the genesis and age of the surfaces are diverse. The background surface type is marine terraces of various (Middle Pleistocene to Holocene) ages, subsequently reworked by fluvial, cryogenic, and other exogenous processes. The classification method used is random forest. For representativeness assessment, we used 211 raster images reflecting the spatial distribution over the area of quantitative characteristics of the earth's surface morphology, as well as substrate parameters (the age of different sediment packs, basement structures), geophysical anomalies, etc. It has been established that using sample data covering fractions of the area percent (0.04-1%) allows for the automatic construction of a geomorphological map for the rest of the area with an accuracy of 78-79%, which is more than 7 times higher than the accuracy of random guessing. The most representative parameters are elevation, various metrics of relative heights/depths, supplemented by texture characteristics of topographic dissection: "non-fractality" and "harmony" of the terrain, etc.
Publication Impact Profile
- Captures
- Mendeley - Readers: 1
Publication details
References11
Breiman L. Random forests. Machine learning. 2001, vol. 45, no. 1, pp. 5�32. DOI: 10.1023/A:1010950718922. https://doi.org/10.1023/a:1010933404324
Bugnicourt P., Guitet S., Santos V. F., Blanc L., Sotta E. D., Barbier N., Couteron P. Using textural analysis for regional landform and landscape mapping, Eastern Guiana Shield. Geomorphology, 2018, vol. 317, pp. 23�44. DOI: 10.1016/j.geomorph.2018.03.017.
Hargrove W. W., Hoffman F. M., Hessburg P. F. Mapcurves: a quantitative method for comparing categorical maps. Journal of Geographical Systems, 2006, vol. 8, no. 2, pp. 187�208. DOI: 10.1007/s10109-006-0025-x.
Kharchenko S.V. Automatic recognition of the landforms origin in the Kola Peninsula based on morphometric variables, Geodesy and Cartography, 2022, vol. 83, no. 2, pp. 12�25. DOI: 10.22389/0016-7126-2022-980-2-12-25. (In Russian).
Kharchenko S.V. New Challenges of Geomorphometry and Automatic Morphological Classifications in Geomorphology. Geomorfologiya, 2020, no. 1, pp. 3�21. (In Russian) DOI: 10.31857/S043542812001006X.
Kozlov E. P., Cherdantsev S. G., Sokolovsky A. P., Novoseltseva R. G., Nikitin Yu. N., Voronov V. N., Suslov S. L. State Geological Map of the Russian Federation. Scale 1: 200,000. West Siberian series. Subseries Tyumen-Salekhardskaya. Sheets R-42-VII-IX, XIII-XV. Explanatory note. Moscow, MF VSEGEI (Publ.), 2015. 103 p. (In Russian)
Maxwell A. E., Shobe C. M. Land-surface parameters for spatial predictive mapping and modeling. Earth-Science Reviews, 2022, vol. 226, pp. 103944. DOI: 10.1016/j.earscirev.2022.103944.
McDermid G. J., Franklin S. E. Spectral, spatial, and geomorphometric variables for the remote sensing of slope processes. Remote Sensing of Environment, 1994, vol. 49, no. 1, pp. 57�71. (In Russian) DOI: 10.1016/0034-4257(94)90059-0.
Mel'nik M.A., Pozdnyakov A.V. Fractals in the erosion dissection and self-oscillations in geomorphosystems' dynamics. Geomorfologiya, 2008, no. 3, pp. 86�95. (In Russian). DOI: 10.15356/0435-4281-2008-3-86-95.
Ng V. W., Breiman L. Bivariate variable selection for classification problem. Technical report. Berkeley: Department of Statistics, University of California-Berkeley, 2005. 22 p.
Open Topography � ALOS World 3d � 30 m. � 2021. � URL: https://portal.opentopography.org/raster?opentopoID=OTALOS.112016.4326.2. Available at 01.01.2022.
View or Download full articleAccess options
SWS access login
Login as SWS Scientific CommitteeLogin as SWS Scientific PartnerLogin as SWS AuthorAuthors and approved SWS contributors will read and export their own linked papers after identity matching by SWS profile, email and SGEM GlobalID.
For librarian assistance: [email protected]
Purchase Instant Access
- Article can be downloaded after successful payment.
- Article may be used according to SWS library access terms.
- Article cannot be redistributed.