The spatiotemporal response of China's vegetation greenness to human socio-economic activities.

Climate change and human socioeconomic activities both strongly impact long-term vegetation greenness. It is more a challenge to evaluate the impacts of socioeconomic activities on vegetative greenness than climate change, partially due to the lack of appropriate quantitative indicators of the former. Here we examined the relationship between the remote sensing nighttime light (NTL) data and the Normalized Difference Vegetation Index (NDVI), which in this study are used as the proxies of socioeconomic activities and vegetation greenness, respectively. We first eliminated the vegetation greenness changes in response to climate change and calculated the human-activities-induced NDVI (HNDVI). After explored the spatiotemporal patterns of the HNDVI and NTL data across China from 1998 to 2018, we studied the relationship between the HNDVI and NTL at the grid and county levels, respectively. Our results show that the mean adjusted DN values of the NTL data (NTLI) continuously increase (+0.2938) across our study area from 1998 to 2018, whereas the HNDVI values fluctuate with a general upward trend (+0.0018). Most grids (91.2%) with increased HNDVI were found in rural areas, particularly in the Northeast forest shelterbelt and the Loess Plateau. By contrast, the HNDVI values in rapidly urbanized areas in Chinese major urban agglomerations mainly show a downward trend, especially in the Yangtze River Delta (YRD) urban agglomeration. The relationships between the NTLI and HNDVI are inconsistent over time and across space, which could be attributed to land use conditions, afforestation projects in rural areas, and greening activities in urban areas over different periods and regions.

Mapping human's digital footprints on the Tibetan Plateau from multi-source geospatial big data.

It is a challenge to accurately quantify short-term dynamic human impact on the environment, which is the key to ecosystem and biodiversity conservation. Human's digital footprints are widely used as a proxy of dynamic human impact. This study developed a method to accurately and objectively map the dynamic human's digital footprints in the Tibetan Plateau using the geospatial big datasets, including the numbers of smartphone location request, microblog check-ins, and geo-tagged flicker photos. We developed a method to calculate the fused digital footprint intensity (FDFI) by integrating the location information in the three datasets. The magnitude of the FDFI was converted to a footprint intensity score (FIS), which was then used to infer the human impact. Results show that the average FIS values in Qinghai and Tibet are low (0.12 and 0.04, respectively). The grids with a positive FIS only account for 5.99% of the Tibetan Plateau and are mainly found in the cities and along the transportation networks. The FIS is also strongly correlated to land use and the positive values are mainly found in the built-up and agricultural lands. All other land use categories tend to have near zero FIS values. We concluded that human activities overall show very limited impact on the Tibetan Plateau and most of the impact is found in the built-up and agricultural lands.