Normalised representation of terrestrial vegetation response to extreme weather events

Since the beginning of the industrial era, the climate of our planet and the human environment have been changing rapidly. Therefore, known and new types of extreme events have been and will continue to be a challenge. An example of a climate challenge is climatological extremes. In recent decades, extreme weather events such as wildfires, floods, droughts, and heat waves have increased across the globe. These extreme events can disturb and alter ecosystems over timescales ranging from minutes to months. However, the recovery and adaptation processes often take far longer than the extreme events. While the intensity of adaptation efforts may vary, they inevitably follow disturbances.

Here, we focus on the recovery dynamics of vegetation in different types of ecosystems after droughts and heat waves, which are the most damaging types of weather extremes. The study covers the last decades period and is based on satellite data.

Disturbance-induced changes in terrestrial ecosystems affect photosynthetic activity, reducing carbon dioxide (CO₂) fixation. We hypothesise that, in return, the temporal dynamics of atmospheric CO₂ fixation by vegetation may indicate different stages of ecosystem recovery - normal ecosystem state (before extreme), imbalance phase, post-extreme phase, recovery phase, or collapse. We find such an approach helpful for understanding the time frames of the phases and capturing phase transitions and general ecosystem states in near real-time.

The identification of vegetation recovery stages is influenced by several factors, including environmental conditions and seasonal cyclicity. To ensure the effectiveness of an automated approach, a unified phase-stage representation for comparability and analysis of CO₂ uptake is required.

To achieve this, we divide daily CO₂ uptake values by their maximum values observed during a year without significant droughts and heatwaves. We have chosen the observation period from 1993 to 2005, which includes a European drought in 2003 and the periods before and after it. As a result of normalisation, stronger ecosystem recovery will correspond to values around “1” and weaker recovery - to a range around “0”. Negative values could indicate the dominance of CO₂ emissions or ecosystem degradation processes.

Vegetation indices, such as NDVI, can be employed as markers of transformation scope to identify the beginning and end of the vegetation growth period activity. This allows us to represent the time scale in a normalised relative interval – [0;1].

The results are a first step towards a normalised representation of the response of terrestrial vegetation to further study the dynamics of its recovery from extreme weather events.