Scholarlink

Abstract <p>Currently, northern ecosystems are changing under the influence of climate warming, which makes it necessary to study their dynamics and predict their further development. The Nadym geocryological observation site landscapes was classified using the Support Vector Machine method with high accuracy (86–97 percent) based on Landsat satellite images from 2001, 2013, and 2023. Temporal sampling with selected landscapes (surface types) made it possible to determine changes in their area and to simulate the development of the area up to 2033 in the MOLUSCE 4.0 QGIS module based on the algorithm of artificial neural networks (multi-layer perceptron), taking into account geophysical influencing factors. It was revealed that the dynamics of landscapes is largely due to the periodic occurrence of fires. Analysis of the transition matrix showed the presence of local substitutions of flat shrub–grass–moss bogs with grass–shrub–moss–lichen peatlands. Forecasting for a 10-year period has demonstrated that, with the current trend, vegetation will recover on forest and peat burns with an increase in lichen abundance and a further smooth transition of a small proportion of bogs to peatlands, due to the processes of drainage and phytomass growth against the background of an increase in the surface atmospheric temperature.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Geoinformation Modeling of Changes in Natural Landscapes in the Southern Cryolithozone of Western Siberia Based on Landsat Satellite Images

  • 期刊论文

  • 发表日期：

    2025年12月1日

查看全文
  • A. V. Fakhretdinov

摘要

Abstract

Currently, northern ecosystems are changing under the influence of climate warming, which makes it necessary to study their dynamics and predict their further development. The Nadym geocryological observation site landscapes was classified using the Support Vector Machine method with high accuracy (86–97 percent) based on Landsat satellite images from 2001, 2013, and 2023. Temporal sampling with selected landscapes (surface types) made it possible to determine changes in their area and to simulate the development of the area up to 2033 in the MOLUSCE 4.0 QGIS module based on the algorithm of artificial neural networks (multi-layer perceptron), taking into account geophysical influencing factors. It was revealed that the dynamics of landscapes is largely due to the periodic occurrence of fires. Analysis of the transition matrix showed the presence of local substitutions of flat shrub–grass–moss bogs with grass–shrub–moss–lichen peatlands. Forecasting for a 10-year period has demonstrated that, with the current trend, vegetation will recover on forest and peat burns with an increase in lichen abundance and a further smooth transition of a small proportion of bogs to peatlands, due to the processes of drainage and phytomass growth against the background of an increase in the surface atmospheric temperature.