| As an important part of the terrestrial ecosystem,vegetation supports the sustainable development of the environment and human society.With global warming and increasing human disturbance,the dynamic changes of vegetation are different among various ecosystems.Scientific and quantitative analysis of the relationship between vegetation,climate and humans is of great significance for formulating strategies to tackle climate change and ultimately achieve the sustainability of regional ecosystems.In this study,we comprehensively analyzed the characteristics of vegetation change trends during 2000-2019 in 34 stations of the Chinese Ecosystem Research Network(CERN),covering five ecosystem types(grassland,desert,agriculture,forest,and wetland).The effect of remote sensing data sources on the evaluation of vegetation change was investigated,with Normalized Difference Vegetation Index(NDVI)as the core indicator.We also evaluated the influence of climate change and human activities on the regional vegetation dynamics with trend analysis,consistency test,Pearson correlation analysis,and redundancy analysis(RDA).The main results were as follows:(1)There was obvious difference among five vegetation indices datasets.The annual average of NDVI?GIMMS were the highest,with NDVI?MODIS and NDVI?SPOT followed.The NDVI value of Landsat is the lowest one.EVI?MODIS was lower than the corresponding NDVI value.This difference could be due to the water vapor absorption capacity caused by the different band bandwidth of the sensor,which results in different sensitivity and saturation in areas with high vegetation coverage or during periods of vigorous vegetation growth.The fluctuations of the vegetation index in different data sets were extremely significantly correlated.Among them,the correlation between NDVI?MODIS and EVI?MODIS was the highest,indicating that the fluctuation correlation of the same sensor was significantly stronger than that of different sensors.It should be noted that NDVI?Landsat had the worst consistency with NDVI from other data sets.It may be related to the limited Landsat data availability and cannot provide stable long-term sequence images.We also found that NDVI?GIMMS with coarse spatial resolution image has a high NDVI value compared to other NDVI data sets.This will become more obvious in extreme weather years,especially for agricultural ecosystems and NDVI?Landsat.Therefore,how to evaluate and select the applicability of different data sources at different scales is still an important issue for the evaluation of vegetation dynamics.(2)A significant upward trend(slope=0.268×10-2yr-1,p<0.01)was observed for the NDVI?MODIS in the CERN during 2000-2019.Specifically,the annual average NDVI of agriculture,forest,and desert ecosystem showed extremely significant upward trends,and the NDVI of wetland and grassland ecosystem showed downward trend.The regions in which average annual NDVI increased rapidly were mainly appeared in the agricultural and forest stations in the middle-lower reaches of Yangtze River and the middle-lower reaches of the Yellow River.The NDVI in the desert and agricultural stations in the west and north relatively increased slowly.Except for the degradation of grassland and wetland in spring,the NDVI of the agriculture,forest,and desert stations all showed a significant upward trend in all seasons.Among them,summer contributed the most to the annual NDVI change.For the 3×3 km buffer zone and administrative districts of CERN station,the NDVI also shown annual growth of0.268×10-2yr-1(p<0.01)during 2000-2019.In general,vegetation index in the CERN station shows an upward trend.However,there were great differences among ecosystem types,geographical scales,and time stages.(3)Climate change and human activities had significant impacts on vegetation dynamics in the CERN.Among them,climate change was the dominant factor.60.1%of NDVI variation was caused by changes in hydrothermal conditions.However,NDVI was not restricted by a single environmental factor but the organic combination of multiple environmental variables.Besides,the differences in NDVI responses between counties and buffer zones also indicated that the larger the geographical area studied,the higher the degree of impact of climate change and human activities on NDVI.Among different types of ecosystems,the dominant factor for grassland NDVI fluctuations was population density.For the NDVI in desert ecosystem,the dominant factors in the desert were precipitation.Agriculture was most affected by temperature changes.Forests were mainly controlled by total radiation and relative humidity.On the whole,climatic factors were the important reason for the variation of NDVI in the CERN ecosystem,while the impact of human activities was relatively limited. |
PDF Full Text Request