[M-IS05] Environmental, Socio-economic, and Climatic Changes in Northern Eurasia

[MIS05-P10] Long-term Land Cover Projections in a Disturbed Peatland Region: A CA-Markov Case Study from Western Siberia

*Vladimir Ivanov¹、Ivan Milyaev^1,2、Evgeniya Soldatova¹ (1.Institute of forest science, Russian academy of science、2.University of Tyumen)

キーワード：Land use and land cover change, GIS, Remote sensing, Drained peatland, CA-Markov model, Landsat imagery

This study forecasts future land cover dynamics in the Tarmanskoe peatland (Tyumen, Western Siberia) using a hybrid Cellular Automata-Markov (CA-Markov) model, with the aim of supporting informed land management and conservation strategies in disturbed peatland regions. Utilizing multi-temporal Landsat imagery (1984–2024), 15 land cover maps generated at 2–3-year intervals provided the basis for model calibration. The CA-Markov approach integrates transition probability matrices with spatial neighborhood influence, enabling the simulation of land cover changes for the years 2034, 2044, and 2054. Model simulations were implemented in Python using GeoPandas, Matplotlib, and NumPy libraries (Harris et al., 2020; Hunter, 2007), and validation against observed changes yielded Kappa coefficients ranging from 0.711 to 0.915, indicating strong predictive performance. The simulation projects a continued expansion of forest ecosystems, particularly small-leaved and mixed forests, on former peat extraction sites. Historically, forests have progressively occupied drained wetland areas since 1990; the model now forecasts that forests will cover approximately 13.7% of the total study area by 2054. In contrast, waterlogged and hayfield ecosystems on former peat extraction sites, which declined sharply from 30% in 1984 to 15.2% in 2024 due to forest encroachment, are predicted to experience only an additional 6.2% reduction over the next three decades, suggesting a slowing trend in meadow loss. Furthermore, the model indicates that residential areas will expand significantly—from a 2.2% increase by 2034 relative to 2024, up to 14.3% of the area by 2054 (a 6.3% increase over the 2024 baseline). In addition, water bodies, which comprised 3.5% of the landscape in 2024, are expected to decline further, retaining only 47.8% of their 2024 extent (or 17.4% of their 1984 extent) by 2054. In contrast, natural wetlands, pine forests, and croplands are projected to remain relatively stable, fluctuating within ±0.5%. Although the Kappa coefficients indicate high model accuracy, we should acknowledge that the CA-Markov model performs better when trends are stable and has limited ability to account for trends that are unstable or driven by anthropogenic pressure and climate change. These projections underscore the dynamic nature of post-extraction landscape succession and highlight the dual challenges of ecological restoration and residential area expansion. The findings provide essential insights for developing targeted management strategies aimed at maintaining hydrological balance, mitigating habitat fragmentation, and preserving ecosystem functions in drained peatland landscapes.
Research was supported by the Russian Science Foundation (No 23-77-10012).
References
Harris, C.R., Millman, K.J., van der Walt, S.J., Gommers, R., Virtanen, P., Cournapeau, D., Wieser, E., Taylor, J., Berg, S., Smith, N.J., Kern, R., Picus, M., Hoyer, S., van Kerkwijk, M.H., Brett, M., Haldane, A., del Río, J.F., Wiebe, M., Peterson, P., Gérard-Marchant, P., Sheppard, K., Reddy, T., Weckesser, W., Abbasi, H., Gohlke, C., Oliphant, T.E., 2020. Array programming with NumPy. Nature 585, 357–362. https://doi.org/10.1038/s41586-020-2649-2
Hunter, J.D., 2007. Matplotlib: A 2D Graphics Environment. Comput. Sci. Eng. 9, 90–95. https://doi.org/10.1109/MCSE.2007.55