| With the rapid development of urbanization,the atmospheric environment is becoming serious,PM2.5 has been widely concerned as an important component of air pollution.Land use/cover types and landscape pattern changes have an important impact on PM2.5 pollution,it is very important to analyze the relationship between landscape pattern and PM2.5 to improve the regional air quality by guiding urban construction.In this study,Zhejiang Province was selected to analyze the influence of landscape index on the PM2.5 concentration.PM2.5 monitoring obervations in 47 monitoring stations and the data of land use landscape pattern around the site over Zhejiang Province were collected.We investigate the effect of urban landscape pattern on PM2.5 spatiaotemporal variability by using both LUR model and GIS spatial analysis.The statistical relationship between landscape pattern and PM2.5 concentration was evaluated firstly.The results showed that:?1?The highest seasonal average PM2.5 appeared in winter,while the lowest value appeared in summer;Linping which located in the urban area of Hangzhou had the maximum annual value of 63.952?g·m-3,while Qiandao Lake had the lowest value of 33.191?g·m-3;?2?Among the five landscape metrics,landscape composition metric?PLAND?and landscape configuration metric?ED?had obvious effects on PM2.5.The resiPLAND5 was the highest,the value was 0.633;?3?ResiPLAND played the most important role in PM2.5 concentration in all regression model.The composition and configuration of construction land?resiED,resiPD,commPD and commPLAND?also appeared in the seasonal regression;The landscape metrics of water,road,and forest also play roles in PM2.5 concentration to some extent.The seasonal Land Use Regression models were established using geographically weighted regression method.The R2 and adjust R2 value were significantly improved when the landscape pattern factors were included in the models.The validation results showed that the prediction error was reduced obviously.The prediction model can explain the distribution characteristics of PM2.5 concentration better in space.Except spring,the other three seasonal models can explain the variation of PM2.5 concentration above 77%,The R2 of the test model was 0.77?spring?,0.79?summer?,0.52?Autumn?,and winter?0.71?.The spatial distribution of PM2.5 concentration in the four seasons is in line with the actual situation,and the middle part of Hangzhou is high and the southwest is low.The results show that the landscape pattern has an impact on the spatial and temporal distribution of PM2.5.It has some reference value and significance for further study of regional PM2.5 estimation.Inconclusion,the land use type and landscape pattern metrics are important factors affecting PM2.5 concentration and its spatial and temporal variability.The results can provide theoretical and practical basis for the PM2.5 pollution controlling,land use planning and urban ecological construction in Zhejiang province and the related eastern part of China. |
PDF Full Text Request