EGU General Assembly 2023
© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

The use of random forest to identify climate and human interference on vegetation NDVI changes in Southwest China

Yuyi Wang1,2 and xi chen1,2
Yuyi Wang and xi chen
  • 1Institute of Surface-Earth System Science, School of Earth System Science, Tianjin University, Tianjin, China
  • 2Tianjin Key Laboratory of Earth Critical Zone Science and Sustainable Development in Bohai Rim, China

Identifying vegetation changes and the associated driving forces provides a valuable reference for developing ecological restoration strategies. However, it remains a challenge to disentangle the impacts of climate, vegetation, and human interference impacts on vegetation changes. In this study, the temporal variations of the Normalized Difference Vegetation Index (NDVI) during 2000 ~ 2015 in space were used to identify the greening (restoration) and browning (degradation) areas in southwest China. The Random Forest (RF) approach was applied to distinguish the main driving forces of the greening and browning areas. Results showed that the RF approach can be effectively used to learn the complex non–linear interactions between vegetation change, local climate, and human interferences. Vegetation greening was prominent in 85.90% of the study area, while 5.59% of the area still experienced significant vegetation degradation. Population pressure was an important factor to alter the sign of long-term vegetation trends. The greening trends are mainly observed in the high elevation areas with low population density (e.g., population density lower than 180 people/km2 and altitudeabove 1000m), which are attributed to both artificial reforestation measures and climate warming. In contrast, the browning trend was concentrated in the low elevation areas with high and temporally intensified population density due to urbanization with a high population density (over 1000 people/km2) and an increased rate (over 20 people/km2 per year). The results of this study strengthen our understanding of the complex convolutions among climate, human activities, and vegetation in southwest China.

How to cite: Wang, Y. and chen, X.: The use of random forest to identify climate and human interference on vegetation NDVI changes in Southwest China, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-2315,, 2023.