Relationship Between POI Spatial Density And Spatiotemporal Changes In Forest Landscape Pattern In Urban Central Area

Urban forest is an important part of urban ecosystem.Studying the dynamic changes of urban forest landscape pattern is conducive to rational planning of urban forest construction and increasing the maximization of urban forest benefits,as well as providing a new perspective of optimization to serve urban residents at a higher frequency.As a key point in the construction of important central coastal cities in China,the study of urban forest has important guiding significance for promoting ecological environmental protection and restoration in coastal zone cities in Qingdao.The study takes the central urban area of Qingdao as the research object,uses POI data,Landsat 7 TM C2 L2 and Landsat 8TM satellite images as data sources,and analyzes the spatial distribution change characteristics of POI data in the central urban area of Qingdao during 2010-2020;explores the spatial and temporal dynamic changes of forest landscape in the central urban area of Qingdao;by constructing a spatial and temporal geographically weighted regression model The spatial heterogeneity between the spatial distribution density of POI and forest landscape pattern in the study area was explored.In addition,the effects of POI density,population density and road network density on the urban forest landscape pattern were compared and analyzed based on multi-source heterogeneous data.In order to comprehensively provide a new perspective of optimization scheme and certain theoretical basis for maximizing urban ecological benefits in Qingdao,and provide a reference for the future urban planning as well as ecological environment construction in Qingdao.The main work and conclusions are as follows:(1)The spatial distribution change characteristics of POI data in the central city of Qingdao were analyzed in detail.The POIs in the study area were spatially analyzed based on the kernel density analysis tool and spatial analysis tool in Arc GIS10.6 software using up to 10 years of POIs as the data source.The results showed that the number of POIs in the central city of Qingdao during 2010-2020 showed an increasing trend year by year;the spatial distribution characteristics of POIs were obvious with great regional variability,concentrated in the core areas of Shinnan District and Shibei District,and the spatial distribution of POIs in Licang District and Laoshan District was more dispersed.The area of the high value of the overall POI nuclear density is gradually increasing.(2)The dynamic changes of forest landscape pattern in the central city of Qingdao were investigated.Based on the forest classification results of Landsat series images,the spatial and temporal dynamic changes of forest landscape from 2010 to 2020 were analyzed from the perspectives of overall forest landscape spatial and temporal distribution,forest area,forest cover and forest landscape pattern.The results show that the forested area in the central urban area of Qingdao city shows a trend of decrease between 2010 and 2020,with a total decrease of12.45 km2;the forested landscape is mainly distributed in the area with elevation class 1(0-200),accounting for 71.63% of the study area,and the area with slope class 2(5-15°),accounting for 34.90% of the study area.During 2010-2020,the largest proportion of forest cover in the central city of Qingdao is 70～100%.On the whole,the forest landscape pattern in the study area basically shows a trend of gradual fragmentation,gradual dispersion,and gradual uniformity,and the fragility of forest ecosystems gradually increases.(3)The spatial heterogeneity between the spatial distribution density of POI and forest landscape pattern was investigated.The spatial autocorrelation of POI density and forest landscape pattern index was calculated,and the spatial and temporal heterogeneity between POI density and forest landscape pattern index in the central city of Qingdao was investigated based on a spatial-temporal geographically weighted regression model.The results showed that POI distribution density and forest landscape pattern index were spatially autocorrelated,and POI distribution density was significantly and positively correlated with ED(edge density),SHDI(Shannon diversity index),NP(number of patches),AWMPFD(area-weighted average patch fractal index),LPI(maximum patch index),and CA(patch area),and PSSD(patch area standard deviation),AWMSI(area-weighted mean shape index),and MSI(mean shape index)were negatively correlated,indicating that the higher the number of POI points,the stronger the disturbance to the forest landscape,the higher the degree of landscape fragmentation,and the more complex the shape of forest landscape patches.(4)The variability of the effects of POI density,population density and road density on the forest landscape pattern in the central city of Qingdao was explored.Based on multi-source heterogeneous data sources,the spatial heterogeneity of POI density,population density and road density with forest landscape index in 2010 and 2020 was analyzed by using a spatiotemporal geographically weighted regression model.The results showed that POI density,population density and road network density showed obvious spatial heterogeneity with forest landscape pattern,and showed homogeneity of positive and negative relationships.Meanwhile,POI density has the greatest influence on AWMPFD,LPI,and ED,road density has the greatest influence on AWMSI and NP,and POI,population,and road density have similar influence on SHDI.