| In recent years,due to the gradual deterioration of the global ecological environment,the relationship between vegetation change characteristics and climate change has become a research focus in the field of environment.Based on the MODIS-EVI data and climate data,this paper used the Mann-Kendall test,Hurst index,the Partial Least-Squares Analysis Method and the Spatial Partial Least-Squares Autoregression Method and other research methods to explore the temporal and spatial variation characteristics of EVI and meteorological factors of vegetation in the sample areas of southwest and northeast China from 2000 to 2019 and the response of EVI of vegetation to meteorological factors.The main conclusions are as follows:(1)In terms of temporal variations,EVI of vegetation in the southwest sample area and the northeast sample area in summer was between 0.37-0.44 and 0.43-0.51,respectively,and the temporal dynamic change of precipitation in the two sample areas was greater than the temperature.Spatially,the vegetation coverage in the sample area of southwest increases from west to east,while the precipitation and temperature decrease from northwest to southeast.The average EVI of vegetation in the sample area of northeast China is about 0.5.The high vegetation cover areas are mainly located in the north and south,and the low vegetation areas are mainly distributed in the northeast plain.In terms of the variation trend and trend sustainability of vegetation cover,76.1%of the southwest sample areas were improved and 9% were degraded.In the sample areas of northeast China,63% were improved and 5.2% were degraded.The change trend of vegetation cover in most areas of the sample areas of southwest and northeast China was negative and continuous.(2)SPLSAR was used to obtain the spatial correlation coefficients of vegetation EVI in the southwest sample area and the northeast sample area were 0.15 and 0.13,respectively;PLS was used to obtain the MSE and variance were 0.68 and 0.62,respectively,in the southwest sample area,while SPLSAR was used to obtain the spatial correlation coefficients were 0.26 and 0.30,respectively.MSE and variance were 0.26 and 0.42 respectively using PLS in the northeast sample region,and 0.23 and 0.25 respectively using SPLSAR.The variance of the regression coefficients calculated using SPLSAR for both regions was smaller than that using PLS.Therefore,SPLSAR,which considers the spatial autocorrelation between response variables,is more scientific and persuasive to analyze the correlation between vegetation coverage and meteorological factors.When the spatial autocorrelation between response variables is stronger,SPLSAR is more reliable for the correlation between vegetation cover and meteorological factors.(3)SPLSAR was used to find that the temperature in the sample area of southwest China and the precipitation in the sample area of northeast China had significant explanatory significance to EVI of vegetation in their respective research areas.Precipitation in the sample area of northeast China had a greater influence on EVI than temperature,and precipitation had no explanatory significance to EVI in the sample area of southwest China.In terms of time,the importance of precipitation to vegetation in the southwest sample area was decreasing,while the importance of temperature to vegetation was gradually stable and kept high.Both precipitation and air temperature play an important role in vegetation in the sample area of northeast China.Spatially,precipitation in most areas of southwest sample area has a negative impact on vegetation,while temperature has a positive impact.The precipitation in the southwest of the northeast sample area has a positive effect on vegetation,while the temperature has a negative effect. |
