A Dynamics Trend Analysis Method of Thermokarst Lakes Based on the Machine Learning Algorithm

The thermokarst lake is one of the most typical thermal and thawing disasters, and also an key sign of permafrost degradation. It has a strong impact on the study of global climate change. In this paper, the Beilu river basin in Qinghai Tibet Plateau was selected as an example. With the global avail...

Full description

Saved in:
Bibliographic Details
Main Authors: Chen, Hong, Liqiang, Tong, Zhaocheng, Guo, Jienan, Tu, Hua, Wu, Peng, He
Format: Conference Proceeding
Language:English
Subjects:
Climate change
dynamic change
Indexes
Lakes
Machine learning
Rivers
thermokarst lakes
Time series analysis
trend analysis
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The thermokarst lake is one of the most typical thermal and thawing disasters, and also an key sign of permafrost degradation. It has a strong impact on the study of global climate change. In this paper, the Beilu river basin in Qinghai Tibet Plateau was selected as an example. With the global availability Landsat data (TM, ETM+, OLI), we obtained the multi-spectral indices, which is closely related to the state of thermokarst lakes rich area. Then, the longterm change trend parameter sets of the multi-spectral indices from 2000 to 2020 are taken as the input data sets of machine learning method to accurately characterize the change state of the thermokarst lakes. Based on the proposed machine learning method, the dynamic change results of the thermokarst lake rich area were obtained pixel by pixel. The results show that it is an effective way for thermokarst lake dynamics analysis within the permafrost region. It is not only helpful to predict and control the change of thermokarst lakes, but also has important practical significance for the study of global climate change.
ISSN:2153-7003
DOI:10.1109/IGARSS47720.2021.9554435