Long-term Extraction And Spatio-temporal Variation Analysis Of Wetlands In The Yellow River Basin Based On Google Earth Engine

The distribution of wetlands in the Yellow River Basin faces challenges of unevenness,and the problems of ecological environment caused by the shortage of wetland resources are increasingly serious.Accurately investigating long-term information on wetlands can contribute to wetland protection,management,and planning.However,the limits of stand-alone operation of big-data calculations for remote sensing in the past have meant that no specific dynamic monitoring study has been conducted on the long-term wetland change in the Yellow River Basin with moderate resolution and annual time scale.Thus,in this study,a new combined extraction rule was proposed,which can efficiently eliminate the interference of shadows and dark pixels.Based on this method,an entire annual-scale wetland dataset in the entire Yellow River Basin for 1986–2020 was built by using all available Landsat images.For the first time,the spatial distribution,change trends,conversion processes,heterogeneity,and driving factors of wetlands in the Yellow River Basin were analyzed exhaustively and comprehensively based on this dataset.Therefore,this study is expected to provide a data support and scientific basis for regional wetland protection and management.The result showed that,firstly,the water extraction results were obtained with an overall accuracy of 99.70%and a Kappa coefficient of 0.90.On the whole,the wetlands in the Yellow River Basin showed a trend of initially decreased(1986–2000)and then increased(2001–2020);however only the permanent water bodies were increasing in most areas while the majority of seasonal water bodies were decreasing.Therefore,during the 35-year change processes,there was a net loss of about 1864 km~2 of seasonal water bodies but a net increase of about 994 km~2 of permanent water bodies.However,as time went by,the loss of the conversion from seasonal water bodies to non-water bodies was slowly decreasing,while the conversion from non-water bodies to seasonal or permanent was gradually increasing.During three periods(1986–2000,2001–2010,and 2011–2020),the bulk of the permanent waters showed a pattern of“decreasing in the first period,then increasing in the second period,and then continuing to increase in the third period,”—a pattern of“decreasing-increasing-increasing”,but almost all seasonal waters shrank in two periods and some even in all periods.In regions dominated by human-made wetlands,non-water areas were substantially converted to seasonal and permanent water bodies;however,in areas dominated by natural wetlands,many permanent and seasonal water bodies were gradually being converted to non-water.Thus,most of the increased water bodies were in the form of artificial lakes and reservoirs,but most of the degraded water bodies were natural wetlands and lakes.Through the analysis of driving factors of wetland area change,this study found that precipitation,temperature,number of hydropower stations,population,GDP,and impervious surface area are positively correlated with the permanent water bodies,while the other five factors except precipitation are negatively correlated with the seasonal water bodies.These six factors were the main impact factors that can explain about 70%of the changes in the permanent water area,but they can only explain about30%of seasonal.It signified that,for seasonal water bodies,they had certain influence over the past 35 years,but were not the main driving factors.