On The Cool Island Effect And Its Impact Factors Of Urban Forest

Urbanization changes the thermal characteristic of the urban surface,intensifies and changes the spatial and temporal distribution of the urban heat island effect.China is undergoing rapid urbanization,so the urban heat island effect is an inevitable problem.Urban forest can decrease the urban surface temperature and thus mitigate the urban heat island effect by forming cool islands.Research on the factors influencing the cool island effect of urban forest and the spatial and temporal changes,and analysis of the quantitative relationship between the impact factors and urban cool island effect will help improve the urban thermal environment.The research,referring to important ecological theoretical ideas and applying remote sensing and geographic information science to field sample surveys,analyzes the intensity,scope,temporal and spatial distribution of urban cool island effect in the central urban areas of Changzhou.The specific research content is as follows.(1)Feature Extracting Methord is used to extract the urban forest information from Landsat 8 remote sensing image,and field sample method is used to survey urban forest structure.The spatial distribution characteristics and landscape structure of urban forests are analyzed.(2)The internal temperature changes of parks are used to represent the intensity of cool island effect,which can reduce external environmental disturbances and improve the evaluation accuracy.RTE method is employed to retrieve the Land Surface Temperature(LST).The horizontal impact factors,namely,area,LST,water coverage,green land coverage,impervious surface coverage,Normalized Difference Vegetation Index(NDVI),Temperature Decrease Amplituted(TDA),Temperature Decrease Range(TDR),and Landscape Shape Index(LSI)of eighteen parks are evaluated.Pearson correlation analysis and regression analysis are then employed for a quantitative analysis of the relationship between urban cool island effect and the horizontal impact factors.(3)Sample plot survey and i-Tree Eco methods are used to collect research data.Leaf area,carbon sequestration,Diameter at Breast Height(DBH),tree height,tree density in the researched area are measured as vertical impact factors.Pearson correlation analysis and regression analysis are then applied to a quantitative analysis of the relationship between urban cool island effect and the vertical impact factors.(4)The Landsat 8 remote sensing images of research area in 2014 and 2016 are used to analyze the relationship between land use and land cover changes and urban cool island effect,and the four remote sensing images from 2016 to 2017,which respectively represent the four seasons of spring,summer,autumn and winter,are used to study the spatial distribution of cool island and heat island.The main conclusions are elucidated as below.1.The urban forest characteristics of the researched area:The total area of urban forest in the researched area is 36.2 Km2,with 679,100 trees in the area,and the Greenland cover rate is 27.8%.In the total area of the urban forest,street tree forest covers 12.17 Km2(including 160,100 trees),affiliated garden forest 14.69 Km2(including 260,000 trees),and scenic recreation forest 9.31 Km2(including 259,000 trees).DBH is mainly distributed from 20cm to 30cm,height distributed from 5m to 10m,and crown width distributed from 6m to 7m.The denser urban forest is scenic recreation forest,the density of which is 277 trees/hectare,and the taller urban forest is the affiliated garden forest,the average height of which is 7.78 meters.As is found in sample plot investigation,the dominant species of street tree forest in descending order are Cinnamomum camphora(60.8%)>Koelreuteria paniculata(16.9%)>Platanus orientalis(5.8%)>Ginkgo biloba(4.8%)>Magnolia grandiflora(2.5%),the dominant tree species is Cinnamomum camphora and the numbers of trees in different species are extremely unevenly distributed.The dominant species of affiliated garden forest in descending order are Cinnamomum camphora(31.3%)>Magnolia grandiflora(16.3%)>Osmanthus americanus(16.3%)>Ginkgo biloba(4.8%)>Trachycarpus fortunei(16.3%)>Metasequoia glyptostroboides(16.3%)>Salix babylonica(16.3%)>Magnolia denudata(1 6.3%).The dominant species of scenic recreation forest in descending order are Cinnamomum camphora(8.78%)>Osmanthus americanus(8.43%)>Acerpalmatum(7.43%)>Metasequoia glyptostroboides(6.1%)>Ginkgo biloba(5.8%)>Trachycarpus fortune(5.4%)>Lagerstroemia indica(4.63%)>Acer buergerianum(4.23%)>Celtis sinensis(4.29%)>Magnolia grandiflora(4.22%)>Photinia davidsoniae(3.87%).The total number of urban forest in the researched area is small,with only 2014 patches.The patch size is mostly 0-0.2 hectare,the patch density is 15.5(patchs/Km2),and the patch fragmentation is high but connectivity is weak.2.The relationship between urban forest cool island effect and horizontal impact factors:Land Surface Temperature(LST),Temperature Decrease Amplitude(TDA),and Temperature Decrease Range(TDR)are used to represent the park cool island effect.LST is negatively correlated with water coverage and NDVI(P<0.05),positively correlated with impervious surface coverage(P<0.05),and has no significant correlation with park size,green land coverage rate,landscape shape index.TDA and TDR are positively correlated with park size(P<0.05),negatively correlated with LST(P<0.05),and have no significant correlation with water coverage,green land coverage,impervious surface coverage and NDVI.Through regression analysis,for maximum thermal environment benefit of urban forests,the threshold of patch size is 1.34 to 19 hectares.Within the threshold,the thermal mitigating function of urban forest patches is strong,while beyond the threshold the mitigating function is obviously reduced.For example,each hectare's increase in patch size leads to a decrease in LST by 0.12?,and when the patch size is over 19 hectares,each hectare's increase in patch size leads to a decrease in LST by 0.03?.When the patch size is less than 1.34 hectares,there is no TDR effect.When the patch size is over 1.34 hectares but no larger than 19 hectares,each hectare's increase in the patch size leads to a decrease in LST by 0.12?.When the patch size is over 19,each hectare's increase in the patch size leads to a decrease in LST by 0.03?,10%increase of Impervious surface coverage leads to an increase in LST by 0.47?,and 10%increase of water coverage leads to a decrease in LST by 0.72?.An increase of NDVI by 0.1 leads to a decrease of LST by 0.68?.An increase of LSI by 1 leads to a decrease of TDR by 0.91 meter,and TDA by 0.22?.3.The relationship between urban forest cool island effect and vertical impact factors:LST has significant negatively correlation with leaf area index(P<0.05),carbon sequestration(P<0.05),and density(P<0.01);TDA has significant positively correlation with leaf area index(P<0.05),but no significant correlation with tree height,DBH.Through regression analysis,it is found that an increase of leaf area index by 1 leads to a decrease of LST by 0.48? and an increase of TDA by 0.46?.Tree density has a significant effect on LST.100 trees' increase per hectare leads to a decrease of LST by 0.7?.10%increase of urban forest carbon sequestration leads to a decrease of LST by 0.88?.4.Temporal and spatial variation of cool island effect:The surface temperature of the downtown area of Changzhou increased from 2014 to 2016.The cool island effect shows significant correlation with urbanization and urban planning.As the process of urbanization accelerates the heat island effect leads to a decrease in distribution of cool island effect and intensity.In terms of spatial distribution,the number of urban heat islands in the southwest and northwest of the researched area has increased because the original cool island area has been transformed into heat island area.The cool island area has increased in the central and northeast of the area,while the heat island area has decreased.In terms of seasonal changes,firstly,the distribution of cool island effect varies in different seasons:Autumn>Winter>Spring>Summer.The cool island effect covers the largest area in autumn,followed by winter and spring,and the cool island effect has the smallest spatial distribution in summer.Secondly,the intensity of cool island effect varies in different seasons:Winter>Autumn>Summer>Spring.The cool island effect is highest in winter,followed by autumn and summer,and is the weakest in spring.In the high temperature seasons(summer and autumn),the horizontal and vertical impact factors have a significant correlation with land surface temperature,while in the low temperature seasons(winter and spring),there is no significant correlation between them.Therefore,on the one hand,it is necessary to reasonably optimize and adjust the urban forest structure to mitigate urban heat island effect.On the other hand,it is necessary to strengthen the cool island effect by increasing the surface water coverage and constructing ventilation corridors,which can both mitigate the urban heat island effect.Clarifying the relationship between urban forests and their cool island effect and proposing effective measures of improving urban thermal environment,this research has important guiding value for urban forest planning.