Applying Geodetector to disentangle the contributions of natural and anthropogenic factors to NDVI variations in the middle reaches of the Heihe River Basin

• Published in: Ecological indicators Vol. 117; p. 106545
• Main Authors: Zhu, Lijun, Meng, Jijun, Zhu, Likai
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01-10-2020
• Subjects: Anthropogenic factors; Geodetector; Natural factors; NDVI variations; The middle reaches of the Heihe River Basin; Geodetector; The middle reaches of the Heihe River Basin; NDVI variations; Natural factors; Anthropogenic factors
• ISSN: 1470-160X; 1872-7034
• DOI: 10.1016/j.ecolind.2020.106545

•Natural and anthropogenic factors interact to influence NDVI changes.•Land use conversion made the greatest contribution to NDVI changes.•Precipitation was the most important natural influencing factor of vegetation growth.•The interactions among factors often enhanced the effect of a single factor.•Factors often influenced NDVI changes in a non-linear way.
The detection and attribution of vegetation changes is a prerequisite for vegetation restoration and management. In arid and semi-arid areas, natural and anthropogenic factors interact to influence vegetation change, making it challenging to disentangle the contributions of driving forces. Here we used NDVI as an indicator of vegetation condition and analyzed its spatial and temporal changes in the middle reaches of the Heihe River Basin from 2000 to 2015. Then we applied the Geodetector method, a robust spatial statistics approach, to quantify the effects of natural and anthropogenic factors on NDVI changes. NDVI across the study area showed a significant increasing trend from 2000 to 2015. Both natural and anthropogenic factors were identified as significant driving forces of NDVI change, and the factors, land use conversion type, mean annual precipitation and soil type, caused the greatest influence. The explanatory power of a single factor was often enhanced when it interacted with other factors. We also found that influencing factors often correlated with NDVI changes in a non-linear way. Our research highlights that the Geodetector method is an effective way to disentangle the complicated driving factors of vegetation change, and our results is useful for projecting vegetation change under future environmental change and taking measures to prevent and mitigate land degradation in drylands.