Characteristics of multi-temporal scale variation of vegetation coverage in the Circum Bohai Bay Region, 1999–2009

Abstract

Long-term spatial–temporal dynamics of vegetation coverage is a key problem of issues include global climate change study, regional ecological process monitoring and ecosystem management. Based on SPOT-VGT 10-day composite data over the Circum Bohai Bay Region from 1999 to 2009, this paper performs the Mann–Kendall test and calculates the trend Slope (β) and Hurst index of time series data to study the temporal trends and long-range dependence of Normalized Difference Vegetation Index (NDVI) on 1km spatial scale, and plots monthly calendar and seasonal succession map using spatial analysis techniques, and then analyze and reveal the spatial–temporal characteristics of vegetation coverage. Main findings are as follows: (1) dense vegetation coverage mainly distribute in mountainous and hilly areas dominated by forest and shrub, followed by vast plain areas with well developed agricultural industry, and low vegetation coverage mainly distribute along part of coastal beaches, surrounding most of urban areas and large inland water bodies and in the semi-arid agro-pasture intertwined zone at the north-west corner of Hebei province. (2) Overall, vegetation coverage is high in summer–autumn and low in winter–spring and the increasing trend from 1999 to 2009 on monthly, seasonal and annual scales are significant in most areas. An increased seasonal difference is detected because the improving trend of vegetation coverage in growing seasons is more remarkable than that in non-growing seasons. Furthermore, significant long-range dependence of NDVI time series is detected in most areas which indicate that the trend of vegetation coverage change from 1999 to 2009 will definitely persist in the near future. (3) Both variation of vegetation coverage in the past and its long-range dependence show significant spatial differences on macro spatial scale, in detail, improving trend of vegetation coverage in Liaoning province is very significant and it will persist in the near future because more than 85% of the pixels in Liaoning province has positive Slope (β) and Hurst index great than 0.5, correspondingly, only very mild signal of enlarged seasonal difference is detected in Liaoning province. On the contrary, very large extent of vegetation coverage in Beijing–Tianjin–Hebei area and Shandong province show significant degrading trend and enlarged seasonal difference despite the overall improving trend in these areas in the past 11years. Areas with degrading trend of vegetation coverage in the past and future trends point to continuing degrading account for 6.71% and 9.84% in Beijing–Tianjin–Hebei area and Shandong province, respectively, while it is below 6.00% in Liaoning province. (4) Areas with degrading trend of vegetation coverage in the past and future trends point to continuing degrading show the overall spatial pattern of scattering on macro scale and aggregating on fine scales, specifically, areas surrounding large inland water bodies, along part of coastal beaches and surrounding urban areas indicates strong tendency of degrading in the past and prospective degrading trend in the near future. (5) Complicated driving forces of vegetation coverage change on macro spatial scale are revealed by the spatial patterns of the coupling index between Slope (β) and Hurst index, in detail, rapid and extensive urbanization undergone on multi-spatial scales is the most important man-made force of vegetation coverage degrading, large scale agricultural activities overall improve the vegetation coverage and enlarge its seasonal difference; while factors include water level changes near large inland water bodies, water–salt dynamics at coastal beach and so on are the main natural driving forces of vegetation degradation. The findings in this paper are useful for wise regional ecosystem management in Circum Bohai Bay Region.