Changes in global vegetation distribution and composition under idealized overshoot scenarios

Within the framework of the project IMPOSE (Emit now, mitigate later? IMPlications of temperature OverShoots for the Earth system) six idealized emission-overshoot simulations have been performed with the Earth System Model NorESM2-LM2 and used as forcing for the 2^nd generation dynamic global vegetation model LPJ-GUESS to investigate the impact of different CO₂ overshoots on global vegetation.

The simulations describe a set of scenarios with high, medium, and low cumulative CO₂ emissions, each of which has a short (immediate) and a long (100 years) peak of cumulative CO₂ emissions before declining towards a baseline simulation where a cumulative 1500 PgC is emitted within the first 100 years. The simulations have been performed in a “World without humans”, i.e. without land-use change, urban areas, fire-suppression, etc. to eliminate the somewhat arbitrary human factor.

The results clearly show that the height of the overshoot has a large impact on global vegetation distribution and composition while its duration does not seem to play a significant role. Overall, we can state that any overshoot scenario results in vegetation patterns that are different from (though converging towards) the non-overshoot baseline simulation. The higher the overshoot, the larger the initial deviation.

We have observed that there is less savannah after an overshoot and less so, the higher the overshoot, due to a reduced amount of tropical rain-green trees. As a result, there is also significantly less potential for fire. Further, there is more boreal vegetation, partly at the expense of temperate summer-green trees. A convergence towards the baseline simulation seems to be possible but isn’t reached by the end of the simulation window.

Furthermore, it can be observed that overshoots are asymmetrical when it comes to succession, i.e. while there are well-known succession patterns when global temperatures rise and vegetation is expanding into previously colder regions, patterns are different when the temperatures on the decline.

Finally, we like to state that dynamic vegetation is an important feature in Earth-system models w.r.t. vegetation carbon sequestration. Not only do the biogeophysical feedbacks matter, the total amount of carbon sequestered is about 16% higher than in simulations in which dynamic vegetation was supressed. These, and other feedbacks will be investigated in more detail in the ongoing Horizon Europe projects OptimESM and RESCUE.