Spatial dependence of vegetation recovery after drought events and the spatiotemporal characteristics

Hydrological extremes pose a serious threat to critical functions and services provided by terrestrial ecosystems. Anticipated increases in frequency and severity of droughts due to climate change is expected to negatively impact ecosystem functioning and vegetation health. The ability of vegetation to recover after a drought episode is an important metric of the drought impact. Wide-spread vegetation drought impacts can result in a regional decrease in ecosystem resilience and an overall decline in ecosystem health. Analyzing the post-drought recovery characteristics of vegetation and it spatial connectiveness provides vital information on the vulnerability to future droughts and its ability to deal with reoccurring drought events or multi-year drought events.

However, the spatial dependence of post-drought vegetation recovery i.e. the extent to which events co-occur at multiple locations simultaneously, is largely unknown and unstudied. In our research, we identify the spatial dependence of vegetation recovery after a drought using complex networks and event synchronization (ES). Thereby we aim to explain the underlying mechanisms and patterns which could potentially support recovery forecasting in the future.

Drought events are selected based monthly SPEI and EVI data, where an ecological drought event is defined as an EVI-anomaly coinciding with a meteorological drought. Based on these events, we create networks of drought event (recovery) co-occurrences. With the use of ES the spatial dependence of different stages of vegetation recovery is quantified. Additionally, regions with similar recovery capacity are evaluated to what degree their recovery responses can be explained by temporal or spatial factors like hydro-meteorological and geographical characteristics.

Our first results show that both geographical, climatological and hydro-meteorological factors are significantly different between regions with similar recovery behaviour. This highlights the relevance of using both temporal and spatial factors when studying the resilience of ecosystems after drought impact.