Dynamic Change Of Vegetation Coverage And Its Influence Factors In Fujian Province

Fujian province is located in the subtropical zone,which has good water and heat conditions,and suitable for vegetation growth.Fujian province is the main commodity forest area in China.However,due to the human factors and natural disasters in long history,the vegetation in some areas of Fujian province has been seriously damaged.Vegetation coverage reflects vegetation growth,which is an important index to evaluate the regional ecological environment.Therefore,it is suggested that studying the trend of vegetation cover and its driving factors scientifically in Fujian province will be a great premise on evaluating the ecological environment quality,maintaining the ecological environment function and developing the ecological protection policy in Fujian.At present,the research about the trend and driving factors of regional vegetation coverage is focused on the influence of natural factors such as vegetation coverage change about the weather and topography.However,the research on human factors such as social economy and urbanization are relatively less.In this paper,we use the methods of correlation analysis and linear regression to reveal the complex relationship between vegetation coverage and many influencing factors.Random forest(RF)is a nonparametric model,which can avoid the complicated relationship between the dependent variable and the independent variables effectively during the period of the analyzed process.This model has been widely applied in the fields of ecology and biology The Geographic weighted regression(GWR)can embed the spatial position of the data into the regression coefficients.The local weighted least squares method is used to estimate the coefficients of each point.The coefficients of the spatial position are estimated by the coefficients of the spatial position.The relationship between vegetation coverage and regional influence factors is reflected,but there are few studies on the trend of vegetation coverage change and its driving factors in the evaluation area.In this paper,MODIS NDVI remote sensing images with a resolution of 500m were used in the Fujian province with the highest forest coverage rate.By using the information of vegetation coverage change in Fujian Province from 2000 to 2015,the random forest model was used to analyze the study area.The influence of meteorological,infrastructure,topography and socio-economic factors on vegetation coverage and the main driving factors of vegetation coverage change.Finally,based on the selected main driving factors,the geometric weighted regression model was used to explore the effect of each driving factor on vegetation cover.The impact of changes in the region to assess the ecological environment in Fujian and the government's afforestation policy to provide a scientific basis.The main findings are as follows:(1)Vegetation coverage in Fujian Province increased gradually from 2000 to 2015,and showed different characteristics in different time and space range.Vegetation coverage was the lowest in March,and the vegetation coverage was the highest in March,and the vegetation coverage was the highest in the four seasons.The vegetation coverage in the autumn was the highest in winter and the second and lowest in spring.The vegetation coverage in the growing season was the highest and the coverage of vegetation coverage is basically the same.Through the fitting slope of the model and R2,the annual vegetation coverage change is more consistent with the change of vegetation coverage.(2)The spatial distribution and vegetation coverage change in Fujian Province showed that the vegetation coverage in the southeastern coastal area changed significantly,and the mean annual vegetation coverage in the urban areas was 0.766-0.911,and the size of the vegetation was Nanping City>Sanming City>Longyan City>Ningde City>Zhangzhou City>Fuzhou City>Putian City>Quanzhou City>Xiamen City.(3)The interpretation and correlation of the model were 83.12%and 0.913 respectively,which were superior to the multiple linear regression models.In the process of the stochastic forest model determines the main driving factors affecting vegetation coverage in Fujian Province.According to the importance of the model and the actual situation of vegetation change in Fujian Province,elevation,topography,slope,recent distance to urban area,income of rural residents,annual precipitation,nearest distance to provincial highway,industrial output value Gas pressure,the nearest distance to the river,the cumulative sunshine hours,the level of urbanization,the nearest distance to the national road and other 13 factors is the main driving factor affecting the change of vegetation coverage in Fujian Province.(4)In view of the relatively spatial heterogeneity between vegetation coverage and main driving factors in Fujian Province,Therefore,the geographical weighted regression model is more effective than the multiple linear regression model,and it reflects the vegetation coverage and the main driving factors in Fujian.The effect of different driving factors on vegetation coverage in the same area are different,and vegetation coverage change is the result of man-made interaction with social factors.Among them,the artificial factors have a good effect on soil and water loss control area,and eastern high speed economic development zone in Fujian province Therefore,the relevant departments need to attach great importance to the construction of ecological civilization,improve the implementation of afforestation,return farmland to forests and grassland and so on.Apart from this,we also need to strengthen publicity and education to enhance the people's awareness of environmental protection.