Identifying the drivers of vegetation changes in Inner Mongolia based on residual analysis and Hasse diagram technique

Exploring the effect of climate change and human activities on vegetation is a key requisite for the reconstruction of regional ecological environments. Therefore, based on long-term vegetation GIMMS NDVI data, climate data, and statistical data, the present study applied the Hasse diagram technique and combined the multivariate regression residual analysis to quantitatively analyze the impact of human activities and climate change on vegetation in Inner Mongolia from detail human activities with some innovations. The results showed that (1) NDVI showed an overall increasing trend over the last 39 years, with an abrupt change in 2000; moreover, vegetation growth was better before the abrupt change (PⅠ: 1982–2000) than after it (PⅡ: 2001–2020), with significant downward trends in Xilin Gol and Hulunbuir. (2) Human activities can promote as well as inhibit vegetation, and the promotion effect was larger during 1982–2000 than during 2001–2020, whereas the inhibition effect was larger during 2001–2020. In addition, during PI, vegetation in Inner Mongolia generally experienced promotion by human activities and climate change, while during PII, climate-driven promotion had the strongest effect, followed by human-driven inhibition mainly distributed in Xilin Gol. (3) The result of the Hasse diagram analysis showed that the dominant pathways of human activities affecting most of the cities were economic factors and urbanization during PⅠ and economization during PII.

Methods
(1) To monitor the vegetation, the AVHRR_version 5 (Advanced Very High Resolution Radiometer) NDVI3g datasets were used because of their high quality, with 0.05° and 1-day spatial and temporal resolution, respectively (download from https://developers.google.com/earth-engine/datasets/catalog/NOAA_CDR_AVHRR_NDVI_V5). To reduce the effects of atmospheric and aerosol scattering, we used the maximum value composite (MVC) method to develop a monthly NDVI dataset. The dataset covers the period from 1982 to 2020. In the text, the growing season of vegetation in Inner Mongolia is defined as April–October.

(2) To further explore the anthropogenic factors inhibiting vegetation growth and achieve a detailed stripping of human activities, we adopted the statistical data and the Hasse diagram technique to be combined. The directed Hasse diagram technique (HDT) is an extension of the ISM explanatory structure model, realization based on the theory of partial order, and has been widely used in several fields, such as chemical risk assessment and environmental science, as well as factor ranking of land degradation, to rank the impacts of each factor. The Hasse diagram technique can be used to reflect the correlations of all elements in an ensemble and has shown good performance in the analysis of drivers of land degradation.