Tree-ring responses to extreme climate events as benchmarks for terrestrial dynamic vegetation models
Abstract. Climate extremes can trigger exceptional responses in terrestrial ecosystems, for instance by altering growth or mortality rates. Effects of this kind are often manifested in reductions of the local net primary production (NPP). Investigating a set of European long-term data on annual radial tree growth confirms this pattern: we find that 53% of tree ring width (TRW) indices are below one standard deviation, and up to 16% of the TRW values are below two standard deviations in years with extremely high temperatures and low precipitation. Based on these findings we investigate if climate driven patterns in long-term tree growth data may serve as benchmarks for state-of-the-art dynamic vegetation models such as LPJmL. The model simulates NPP but not explicitly the radial tree ring growth, hence requiring a generic method to ensure an objective comparison. Here we propose an analysis scheme that quantifies the coincidence rate of climate extremes with some biotic responses (here TRW or simulated NPP). We find that the reduction in tree-ring width during drought extremes is lower than the corresponding reduction of simulated NPP. We identify ten extreme years during the 20th century in which both, model and measurements indicate high coincidence rates across Europe. However, we detect substantial regional differences in simulated and observed responses to extreme events. One explanation for this discrepancy could be that the tree-ring data have preferentially been sampled at more climatically stressed sites. The model-data difference is amplified by the fact that dynamic vegetation models are designed to simulate mean ecosystem responses at landscape or regional scale. However, we find that both model-data and measurements display carry-over effects from the previous year. We conclude that using radial tree growth is a good basis for generic model-benchmarks if the data are analyzed by scale-free measures such as coincidence analysis. Our study shows strong reductions in carbon sequestration during extreme years. However, for a better understanding of the impact of extreme events on e.g. the long-term fate of the European carbon balance, more long-term measurement data and improved process-based models are needed.