Extraction And Temporal And Spatial Distribution Characteristics Of Aquatic Vegetation In Typical Lakes In Northern China

Aquatic vegetation plays an important supporting role in the balance of lake ecosystems,and its abundance and cover changes affect the balance of lake ecosystems.Rapid and accurate acquisition of large-scale aquatic vegetation change characteristics and historical spatial distribution can help provide scientific data for lake ecosystem environmental protection policies.The northern lakes in China account for 37% of the total lake area in China and play an important strategic supporting role in ensuring the socio-economic development and ecological security of China.Therefore,it is important to understand and grasp the historical distribution data of aquatic vegetation in northern lakes in China and the trend of change for the ecological management of lakes.Remote sensing methods based on remote sensing data and LUCC research methods play an important role in exploring the spatial distribution of aquatic vegetation,mechanisms and trends of aquatic vegetation change.Therefore,in this study,based on the random forest classification method,CA-Markov model,the spatial distribution of aquatic vegetation in five typical lakes in northern China in 2020,2010,2000,and the 1990 s,1980s,1970 s,and 1960 s and 1950 s were remotely sensed and reconstructed,respectively,and on this basis,combined with structural equation modeling,the spatial and temporal dynamics of aquatic vegetation in typical lakes in northern China were analyzed.On the basis of these simulations,the spatial and temporal dynamics of aquatic vegetation in typical lakes in northern China were analyzed,together with structural equation models,in order to provide a data base and decision support for subsequent studies on the response of aquatic vegetation to climate change.The main work is as follows:（1）Based on multi-period Landsat remote sensing image data,the spatial distribution ranges of aquatic vegetation in five typical lakes in the north in the 1970 s,1980s,1990 s,2000,2010 and 2020 were extracted using the random forest model method,with an average classification accuracy of 85.42% and an overall classification accuracy of over 80%.（2）Based on the above extracted spatial distribution data of aquatic vegetation in six periods in five lakes,the improved results of CA-Markov with model simulation error rate index（MSER）test were proposed,and the spatial distribution of aquatic vegetation in the 1950 s and 1960 s was realized by improving the CA-Markov model,combining different scenarios of meteorological elements changes in different centuries,with 20 and 10 years as steps,respectively.The overall accuracy of the improved CAMarkov model was improved by more than 40% compared with the direct simulation,and the spatial distribution of aquatic vegetation in the historical period was reconstructed with a high degree of confidence.（3）From the 1950 s to 2020,typical lakes in the north show a change trend of aquatic vegetation area decreasing,among which the northeast region shows a change trend of aquatic vegetation area rising and then decreasing;the north China region shows a long-term trend of aquatic vegetation area decreasing;the spatial distribution of aquatic vegetation in lakes in the northwest region decreases slowly;the structural equation model is used to analyze the driving factors of aquatic vegetation change,which shows : In northeast China,industrial pollution and the application of agricultural fertilizers are the main causes of aquatic vegetation change;in north China,population development and economic expansion of human society are the main drivers of regional aquatic vegetation change;in northwest China,the expansion of arable land and the application of agricultural fertilizers are the main influencing factors of aquatic vegetation change.