Heterogeneity, marginality, stagementation and driving forces in the Otindag Sandy Land and its ecotones based on GIMMS NDVI3g v1.0

Abstract

Due to differences in spatial and temporal scales and perspectives, there are academic debates about the understanding of the ecological evolution in Otindag Sandy Land. How vegetation in the sandy hinterland and marginal intersects interacts with the regional climate system and how it responds to human activities remain to be addressed. The motivation for this study is to try to clarify the abovementioned geographical misunderstandings. By setting up a core area (CA), a check area (CK) and four ecotones, various datasets such as the remote sensing vegetation index products GIMMS NDVI3g v1.0 (1982–2015), meteorological data (1982–2015), night-time lighting data (1984–2020), population data (1986–2016) and livestock data (1986–2016) were adopted; and maximum value composite (MVC), average value composites (AVC), linear regression analysis (θslope), phenology analysis (SOS, LOS, and EOS), coefficient of variation (Cov) analysis.and Hurst index (H) prediction were performed; and partial data consistency was validated by MODIS NDVI (2000–2015) dataset. The study found that: (1) vegetation coverage had spatial heterogeneity; (2) vegetation phenology exhibited a spatial differentiation; (3) vegetation growth rate had spatial differences; (4) vegetation response to climate change was regional and significant; (5) the coefficient of variation (Cov) of vegetation was segmental and regional; and(6) the response of vegetation to human activities was not significant. The results demonstrate that the NDVI development in the Otindag Sandy Land and its ecoregions is spatially heterogeneous, regional, marginal and temporally segmental and is more sensitive to climate change than to human activities.