Abstract:
Objectives Vegetation is the main body of terrestrial carbon sink. The relationship between the temporal and spatial variation characteristics of Normalized Difference Vegetation Index (NDVI) and the driving factors are of great significance in the research of regional ecological stability.
Methods Based on the datasets of vegetation NDVI, climate, soil, vegetation type, and topographic, the driving factors and temporal and spatial variation characteristics of vegetation NDVI in Zoige region from 1998 to 2020 were analyzed by using linear regression analysis, coefficient of variation and geographic detector model.
Results The results showed that: 1) Vegetation in Zoige region showed a fluctuating upward trend from 1998 to 2020, with an average annual growth rate of 0.004 3 and an average annual vegetation NDVI of 0.79, showing an obvious trend of vegetation improvement. The overall spatial distribution was high in the southeast and low in the northwest. The proportion of area with annual average vegetation NDVI above 0.7 was 96%. 2) The inter-annual variation rate of vegetation NDVI in Zoige region was low from 1998 to 2020, with an average value of 0.00018, and the area ratio of the stable area was 99.5%. The vegetation status was generally stable in the past 23 years, the average variation coefficient of vegetation NDVI was 16.50%, and the proportion of the area with variation coefficient lower than 20% was 85%. 3) Altitude and annual mean temperature were the dominant natural geographic factors for vegetation NDVI variation in Zoige region, slope and aspect had little effect on vegetation NDVI. There were interactions among the physiographic factors on vegetation NDVI, which were two-factor enhancement (11.11%) and nonlinear enhancement (88.89%) effects.
Conclusions The vegetation of Zoige region has improved in the past 23 years. The changes in natural geographical factors such as elevation and temperature should be fully considered in the selection of vegetation protection and restoration measures and models.