Font Size: a A A

Analysis Of Wetland Dynamics And Drivers In Different Types Of Permafrost Zones In The Greater And Lesser Xing’an Mountains

Posted on:2024-05-02Degree:MasterType:Thesis
Country:ChinaCandidate:Z L HuangFull Text:PDF
GTID:2530306917463494Subject:Cartography and Geographic Information System
Abstract/Summary:PDF Full Text Request
Wetland ecosystems have unique ecological functions and are vital to human survival and sustainable development.According to statistics,the total area of wetlands in China accounts for about 10% of the global wetland area,among which the Great and Less Xing’an Mountains is one of the main distribution areas of wetlands in China.The wetlands in this area are important source areas of rivers for water conservation,as well as valuable habitats for endangered animals,which are of great value in ecological conservation and utilization.According to the research of Chinese scholars,Greater and Lesser Xing’an Mountains are divided into Predominantly Continuous,Predominantly Continuous Isolated,Sparsely Island multi-year Permafrost zones and Seasonally Frozen Ground,and the permafrost degradation in the Greater and Lesser Xing’an Mountains is closely related to wetland degradation.Wetland degradation has triggered a series of frequent ecological problems,so it is important to analyze the changes of wetlands in each permafrost zone of the Greater and Lesser Xing’an Mountains and explore the driving factors of wetland degradation for the conservation and restoration of wetlands in Greater and Lesser Xing’an Mountains and the protection of ecological barriers in the region.This study uses multi-source remote sensing data and remote sensing technology to extract wetland information of each permafrost zone in the Greater and Lesser Xing’an Mountains in 1990,2000,2010 and 2020 on a big data cloud platform.The change characteristics of wetlands in each permafrost zone over 30 years were analyzed through dynamic attitude,transfer matrix,landscape pattern index,mass center analysis and index change,and the drivers of wetland degradation in each permafrost zone in the Greater and Lesser Xing’an Mountains over 30 years were quantitatively analyzed using geo-sensing.The main findings of this study are as follows.(1)In this study,a wetland classification system for the Greater and Lesser Xing’an Mountains was constructed to classify and extract information on the distribution of wetlands in each permafrost zone of the Greater and Lesser Xing’an Mountains.By referring to the Technical Guide to Remote Sensing Classification of Alpine Wetlands,field survey results and relevant literature,a wetland classification system for the Greater and Lesser Xing’an Mountains region was established.The classification system was divided into nine categories: forest,water,farmland,residential,grassland,saline land,bare land,forest wetland and marsh.It also combines multi-source remote sensing data and selects spectral,radar,topographic,textural and index features for classification in order to extract wetland information more accurately.(2)The GEE platform uses multi-source remote sensing data and builds a random forest model to extract wetland information,and calculates a confusion matrix for each classification period to verify the accuracy.The classification results all performed well,with the overall accuracy(OA)of the classification above 85%,the kappa coefficient above 0.83,and the accuracy of producer accuracy(PA)and user accuracy(UA)of the wetland classification above 80%,so that the information of each class and wetland was extracted more accurately in this study.(3)An analysis of the dynamics of wetlands in the permafrost areas of the Greater and Lesser Xing’an Mountains over a 30-year period shows that the total area of wetlands in the Greater and Lesser Xing’an Mountains has decreased by approximately 40% in 30 years.Of these,the area of forested wetlands decreased by 69% and the area of marshes decreased by 32%,with the degradation of forested wetlands in the Greater and Lesser Xing’an Mountains being more severe.The greatest rate and percentage reduction in the area of both types of wetlands in the four permafrost zones is in the Seasonally Frozen Ground.The transfer matrix shows that the two types of wetlands in the Seasonally Frozen Ground of the Greater and Lesser Xing’an Ranges are mainly transferred out to farmland,while the two types of wetlands in the remaining permafrost zones are mainly transferred out to forest.It is also shown from the landscape pattern index that the dispersion of the two types of wetlands increases across the tundra over a 30-year period,with a decreasing trend in the landscape dominance of the two types of wetlands in the region.(4)According to the factor detector,all factors contribute to wetland degradation.Among the permafrost zones,climatic factors(increased air temperature,increased ground temperature and decreased precipitation)and topographic factors(elevation and slope)have a greater impact on wetland degradation,while anthropogenic factors have a smaller impact on wetland degradation.Compared to Predominantly Continuous Permafrost zones and Predominantly Continuous Isolated Permafrost zones,land surface temperature in Sparsely Island Permafrost zones and Seasonally Frozen Ground have a greater effect on wetland degradation.According to the calculation results of the geodetector,the degradation of wetlands in each permafrost area of the Great and Small Xing’an Mountains is mainly due to the combined effect of multiple factors,with climatic factors playing a dominant role.
Keywords/Search Tags:Wetland, Multi-source data, Random Forest, Geodetector, Driving factors
PDF Full Text Request
Related items