Increasing Carbon Cycle Impacts in Response to Extreme Climate Events during the Last Three Decades

Previous studies have shown that climate extremes have strong effects on the carbon cycle and a few negative events account for disproportionally large decreases in global carbon uptake. As the frequency, duration and magnitude of extreme events are projected to increase under a changing climate, quantification of the impact on the carbon cycle becomes even more important.

A related question is if the increasing number of climate extremes in the recent decades has already impacted the carbon cycle? In this study, we analyze the spatiotemporal impacts of climate extremes on gross primary production (GPP) using three global data sets from various sources all covering the last 35 years and driven with consistent climate data. These include upscaled observations (FLUXCOM), satellite-derived light use efficiency, as well as the most recent ensemble of land surface models under the TRENDY protocol (TRENDYv7).

We show results of the sensitivity in the response of the TRENDY model's GPP to extremes in precipitation and temperature compared to the two observation-constrained approaches. Compound events, where both precipitation and temperature anomalies occur simultaneously, are of special interest as they are responsible for the most severe impacts. Our results show that climate-driven GPP extremes are in general responsible for a considerable reduction in the land-carbon sink. The TRENDY model ensemble reveals large differences between the individual models and their response in GPP to climate extremes. In addition, the frequency and magnitude of climate extremes increased over the past decades with increasing negative impacts on the carbon cycle. This study thus provides evidence of a strong positive carbon cycle feedback in association to climate extremes.