| Changes in global surface temperature and precipitation distribution,led by global climate change,profoundly impact wetland ecosystem.As one of the three major terrestrial ecosystems,wetland ecosystem plays an irreplaceable role in ecological value and biodiversity.The Yellow River Basin,an important ecological barrier in northern China,belongs to semi humid and semi-arid areas.Its wetlands play an extremely important role in maintaining regional ecological functions and biodiversity.Wetlands in semi humid and semi-arid areas are sensitive and vulnerable to climate change,and are vulnerable to the dual interference of human factors and environmental factors.The large-scale and long-term monitoring of wetlands has been limited by many conditions such as technology and manpower.Therefore,developed a wetland classification system based on Google Earth engine(GEE)cloud platform in this study,used Landsat images to classify the wetlands in the Yellow River Basin in the past(1990-2020),and analyzed the temporal and spatial variation characteristics of wetlands.Combined with different climate scenarios in the future provided by cmip6,this study built a set of wetland prediction model to realize the simulation of future wetland distribution,then analyzed the prediction results and change characteristics of wetland distribution,and finally put forward protective suggestions for future wetlands.The research work and main conclusions of this paper are as follows:(1)Based on GEE cloud platform and JS programming language,a large-scale and automatic wetland identification and classification system in the Yellow River Basin is designed and developed.GEE cloud platform is a multi-functional cloud platform.It constructs a new multi-dimensional feature classification dataset by introducing texture features,spectral features and terrain features.Using the Random Forest(RF)and Support Vector Machines(SVM)classification algorithm provided by GEE,it performs classification by adjusting segmentation scale and input band,and verifies the classification accuracy through Kappa and overall accuracy(OA).The results show that the RF algorithm is more suitable for the large-scale classification of GEE cloud platform,and the wetland classification of the study area can be realized in combination with the system program.(2)The temporal and spatial variation characteristics of wetlands in the Yellow River Basin were quantitatively analyzed by transfer matrix,wetland dynamics,landscape pattern index and standard deviation ellipse.From 1990 to 2020,the total area of wetlands in the Yellow River Basin showed a trend of first decreasing and then increasing,with a total increase of 1518.47km~2,including 409.16km~2 of rivers,63.92km~2 of lakes,1228.18km~2 of reservoirs,45.15km~2 of snow,120.37km~2 of beaches,997.93km~2 of bottomlands and 867.76km~2 of swamps.From analyzing the landscape pattern index,the continuous increasing patches indicates that the connectivity of wetlands has been destroyed.And the area change trend of the largest patch is not obvious,indicating that the area change of the largest wetland patch is relatively stable.The increasing trend of shape index indicates that the shape of wetland landscape tends to be complex,and the agglomeration index decreases to a low extent,indicating that the agglomeration degree of wetland decreases.The results show that the wetland is exacerbated by human factors,and the overall wetland landscape pattern tends to be complex,fragile and broken.According to the standard deviation ellipse,the wetland change area in the Yellow River Basin shows a decreasing trend,and the wetland change distribution is more concentrated with obvious trend.(3)Quantitative description impacts the importance of factors wetland distribution,and analyze the leading factors affecting the change of wetland temporal and spatial pattern.According to the characteristics of the study area,10 socio-economic data and 6geographical environment data are selected as the influencing factors,and the RF algorithm is used to analyze and mine the factors of various land use expansion and driving forces.The comprehensive ranking of the influencing factors of various wetland types shows that the impact of climate environmental factors(DEM,average annual temperature and average annual precipitation)on Wetlands in the long-time series is greater than that of socio-economic factors(population,GDP and distance to primary roads).(4)Predict the future temporal and spatial pattern of wetlands in the Yellow River Basin under different climate scenarios.With constructed wetland prediction model and adjusted parameters,the wetland distribution data from 1990 to 2005 and other influencing factor data are used to simulate the wetland distribution in 2020.The accuracy of the prediction results is 0.78,indicating that the model has good simulation ability.Using the climate data of SSP1-1.9,SSP 2-4.5 and SSP 5-8.5 released by cmip6,the past and future climate change characteristics of the study area are analyzed.Combined with the wetland prediction model,taking the average annual temperature and average annual precipitation of the Yellow River Basin under different scenario models as replacement variables,nine different climate scenario models are established to predict the future wetland.The results showed that the total area of wetland decreased first and then increased in the future,but it still showed a net increase.The changes of different types of wetlands are limited to the original areas without showing large amount of regression or expansion.(5)Analysis of wetland change types under different climate scenarios in the future.According to the degree of wetland degradation and restoration,the wetland change types are divided into five levels.The results show that the area of obvious restoration is significantly higher than that of slight restoration,and the area of serious degradation is significantly higher than that of slight degradation.The degraded area of wetland is increasing first and then decreasing,and the total restoration area of wetland is rising.In the end,some suggestions are put forward from the aspects of policy formulation and scientific protection.This paper contains 40 figures,24 tables and 193 references. |
PDF Full Text Request