Spatiotemporal Patterns of Vegetation Drought Resilience in China Across Various Climate Zones

Abstract

Under global warming, frequent droughts put more pressure on the productivity of vegetation, so it is crucial to quantify the drought resilience of vegetation to better understand how to make the ecosystem more adaptive to climate change. In this study, we constructed a framework for assessing vegetation drought resilience from three components: hazard, exposure and adaptability, and revealed the spatiotemporal patterns of resilience across various climate zones. The results showed that vegetation resilience in arid regions was better, while vegetation in tropical and temperate regions needed to be paid more attention. In the period 2001-2020, 45.3% of the regions showed a trend of increasing resilience, while 36.4% of the regions had a decline trend, with an overall improvement in resilience. At the same time, we explored the contribution and interaction of the drivers by using the Bayesian-optimized LightGBM model and SHAP algorithm to explain the contribution of factors and their interaction. The results showed that temperature, sunlight hours, potential evapotranspiration and precipitation were the most important four drivers. Increase in sunlight hours has been accompanied by a decrease in precipitation, and the vegetation in China has not been as resilient as it should have been under the combined effect. These results can assist vegetation management in addressing drought-affected vegetation and building a more resilient ecosystem.

Keywords: Vegetation ecosystem, Drought resilience, Climate zones, LightGBM, Interaction