Ecological Indicators (May 2021)
Abiotic disturbances affect forest short-term vegetation cover and phenology in Southwest China
Abstract
Forest disturbances significantly affect the global carbon cycle by, for example, vegetation loss or changing forest phenology. However, the lack of historical disturbance events constitutes a challenge for in-depth temporal and spatial analysis. Available remote sensing time series and combined climate data may have great potential to quickly and consistently detect and identify forest disturbances events. We employed time-series data (2001–2014) of a vegetation index (normalized difference vegetation index, NDVI) and a change detection algorithm (the breaks for additive seasonal and trend, BFAST) to detect forest disturbances in a sub-tropical area located in Southwest China. Remote sensing and meteorological data were combined to distinguish among the typical forest disturbances: fires, extreme cold events in winter (ECE), and droughts. With the reconstructed historical disturbance events, post-disturbance vegetation loss, short-term vegetation cover, and phenology changes were analyzed. Our results show that fires and droughts caused severe damage to forest cover (NDVI anomalies can reach up to −1.84 and −1.11, respectively). Fire changed the regular phenological periods which last 3–4 years, and it also took 1–2 years for vegetation greenness to recover after ECE and droughts, which triggered carbon emissions and reduced forest stocks. Warmer areas were vulnerable to ECE effects as well and should be paid more attention. Post-disturbance effects show complex patterns: characteristics of disturbances, climatic conditions, and multiple events overlaying contribute to modifying forest vegetation. Hence, forest disturbances cannot be neglected but should be emphasized in future forest ecosystem modeling or analyzing. The approach used in the study can be a crucial step in detecting and assessing the effects of various disturbances on forest vegetation and phenology and, thereby, contributes to improved risk analysis and management in forestry.
