Vegetation Dynamics And Its Influential Factors In The Yellow River Basin During 1982 To 2015

Long time dynamic monitoring of the vegetation is an important part in the study of land ecosystem process and global change.Quantitative study on the response of vegetation distribution and changes to various factors is a hot topic in the study of vegetation ecology.Based on the GIMMS NDVI data and multi-source auxiliary data,this paper analyzed the non-linear change trend and stage change characteristics of NDVI in the Yellow River Basin(YRB)from 1982 to 2015 by the method of ensemble empirical mode decomposition.The time lag response of vegetation change to climate factors on the monthly scale and the response of vegetation growth to water-heat status on the annual scale were discussed using the correlation analysis.On this basis,the residual analysis was adopted to obtain the relative contribution rate of human factors and climate factors to the vegetation change in the YRB by removing the climate-dominated trend.Then,the influencing forces were further divided into 12 influencing factors,and the influence of multiple influencing factors on the vegetation distribution and the interaction between the influencing factors were detected by the geographic detector model.The main conclusions were as follows:(1)NDVI in the four seasons and growing seasons of the YRB showed the distribution characteristics of "low in the northwest,high in the southeast and the south".Specifically,the vegetation coverage in the northern piedmont of Qinling Mountain and Ziwuling forest area in the south of the basin were the highest,followed by that in the east of Qingnan plateau in the upper reaches,Qilian Mountain,Guanzhong Plain in the middle reaches,and alluvial plain in the lower reaches.The vegetation coverage in the Loess Plateau and Ordos Plateau in the middle reaches was poor.(2)NDVI in the growing season of the YRB showed a parabolic growth trend,namely,NDVI in the growing season was increasing with an increasing rate.Almost 39.3% of the YRB area experienced NDVI that initially decreased and then increased during the study period,and 20.37% of the area saw NDVI that initially increased and then decreased.Another nearly 39.15% of the region exhibited continuously increasing NDVI,while only about 1% of the region exhibited a continuously decreasing NDVI.(3)The analysis of the increment of NDVI in different periods showed that the proportion of the area with significant increase of NDVI was increasing gradually over time,while the area without significant change of NDVI was decreasing gradually.And the proportion of areas with significant decrease of NDVI in multiple periods were all very small,and the changes were also not obvious.From 2000 to 2010,the proportion of area expericing increase of NDVI obviously increased rapidly.The year of 2000 to 2010 was the period which had seen the most obvious vegetation restoration in the YRB.(4)The temperature of the YRB in March had a greater impact on the growth of vegetation from March to June,while the temperature had a smaller impact on the growth of vegetation in July and August.The precipitation in March inhibited the growth of vegetation in March,and that from May to August generally promoted the vegetation growth,and there existed a one month lag period of vegetation to precipitation.(5)The relative contribution rate of human factors to vegetation restoration was more than 80% in 33.56% of the total area,which was mainly distributed in the north of Yan'an City and Yulin City in the north of Shaanxi Province,the east of Ordos City in the Inner Mongolia Autonomous Region,as well as most of the Loess Plateau in the middle of Gansu Province.These places were also the key implementation areas of ecological engineering.Near the provincial capital cities in the basin,the relative contribution rate of vegetation degradation caused by human factors could reach more than 80%.(6)The effect of each influencing factor on spatial distribution of NDVI was in order of meteorological factor > social economic factor > natural geographical factor.The area of afforestation and output of the second industry,output of the second industry and relative humidity,elevation and relative humidity were three main interaction types affecting NDVI distribution in 2002.The area of afforestation and grain output,output of the second industry and sunshine hours,grain output and sunshine hours were three main interaction types affecting NDVI distribution in 2008.The area of afforestation and grain output,grain output and sunshine hours,the area of afforestation and output of the second industry were three main interaction types affecting NDVI distribution in 2015.