[Spatio-temporal Correlation Between Green Space Landscape Pattern and Carbon Emission in Three Major Coastal Urban Agglomerations].

In order to study the influence of urban green space landscape pattern on urban carbon emissions， nighttime lighting data， socioeconomic development data， and land use remote sensing data from 2000 to 2020 are used as the basis of analysis， and the three major coastal economically developed regions in China-Bohai Rim， Yangtze River Delta （YRD）， and Pearl River Delta （PRD） （nearly 100 cities in total） are used as the study area to analyze the spatial and temporal evolution characteristics of urban carbon emissions， as well as the influence of urban green space landscape pattern and its spatial and temporal changes. We also explored the influence of 10 urban green space landscape pattern indices on urban carbon emissions by using the random forest model and the Lasso regression model and further analyzed the four factors （number of patches， density of patches， dispersion of patches， and complexity of the shape of patches） that had a greater influence by using the spatio-temporal geographically weighted regression model， to explore the results of the spatial and temporal evolution of the influence of the urban green space landscape pattern on carbon emissions. The main findings of this study are as follows： ① Carbon emissions in the three study areas showed a slow growth trend， with the Bohai Rim showing a relatively fast growth rate. Carbon emissions were spatially aggregated in the selected study areas， with the majority of cities in the "high and high" agglomeration and the "low and low" agglomeration regions. There was spatial aggregation of carbon emissions in the selected study areas， with the majority of cities in "high and high" agglomeration and "low and low" agglomeration. The land-averaged carbon emissions in the three study areas were dispersed in all directions， with the economically strong cities as the core， and the overall carbon emission level was dispersed from the center to the surroundings. Additionally， along the rivers and coastal areas， carbon emissions were higher due to the concentration of ports， industrial zones， and cities. ② Landscape occupied by patches， number of patches， and density of patches had a significant negative correlation with urban carbon emissions， which indicates that the higher the number， density， and proportion of the landscape occupied by urban green space patches， the more it could hinder the growth of carbon emissions. On the contrary， the shape index and patch fragmentation index had a positive correlation with urban carbon emissions， indicating that the higher the shape complexity of urban green space patches and the higher the fragmentation degree of patches， the more it promoted the growth of urban carbon emissions. In addition， the aggregation index also showed a significant negative correlation with urban carbon emissions， which indicates that the higher the degree of aggregation of patches， the more it could inhibit the growth of carbon emissions. ③ The correlation between the green space landscape pattern index and carbon emissions showed significant spatial and temporal differences， with large changes around 2010. In the Bohai Rim Region， the influence of the urban landscape pattern index on carbon emissions remained relatively stable， and its influence over time generally showed a decline. In the YRD Region， the shape complexity and dispersion of urban green space had a greater impact on carbon emissions than the number of patches and patch density factors. However， on the contrary， in the PRD Region， the impacts of the number of urban green spaces and density index were increasing. In addition， the spatial influence changes on all showed the clustering of regression coefficients. The impact of urban green space on carbon emissions varied greatly across locations and time， suggesting that policy makers cannot rely on a one-size-fits-all approach to urban green space planning. In the Bohai Rim Region， it is more important to balance the distribution of urban green space with other land uses to maintain stability； in the YRD Region， highly fragmented and overly complex green space patch planning should be reduced； and in the PRD Region， priority should be given to increasing the amount and distribution density of urban green space.