Artificial tree rings from climate simulations - Assessment of uncertainties of temperature reconstructions of the last millennium
Final Report Abstract
This project aimed at an assessment of methods to reconstruct temperatures from tree ring data. Instead of using real tree ring chronologies, the problem is tackled by a simulation approach involving as key element the generation of artificial tree ring sequences. These artificial tree ring sequences were generated by means of two tree ring growth models using climate data from an Earth system simulation of the last millennium. The first tree ring growth model has been obtained from a suitable modification of the dynamic vegetation model LPJ‐GUESS. The second is a simplified (lite) version of the Vaganov‐ Shashkin model of cambial cell growth. In contrast to the real situation of reconstructing past climates, this simulation setup has the advantage that reconstructions of simulated climate from the artificial tree rings can be fully assessed for its reconstruction quality because the simulated “past” climate is fully known. As planned, the model framework for generation and analysis of artificial tree ring sequences has been developed. This development covered the majority of project time. Originally it was intended to use the FORMIND model to generate artificial tree ring sequences, but for technical reasons and because development support could not further be guaranteed by the FORMIND development group at UFZ we had to switch to another tree growth model. With support from our new project partners at BiK‐F we partly re‐programmed the dynamic vegetation model LPJ‐GUESS to trace the growth of individual trees. In addition, we set up scripts to run this modified LPJ‐GUESS (dubbed LPJ‐ST) and VSL over selected geographical regions using climate data from the MPI‐ESM Millennium simulation, and scripts to process the simulation outputs to arrive at tree ring indices and finally at reconstructed temperatures. With the simulation framework at hand, the Millennium simulations have been analyzed for Europe as planned and first investigations of reconstruction methods have been performed. Examples have been given in the report. It has been successfully demonstrated that the framework is now ready for a detailed assessment of tree ring based climate reconstruction methods. We could show that our simulation based approach to assess tree ring based climate reconstruction methods has a large potential. Moreover, thanks to this project, the framework for generation and analysis of artificial tree ring sequences is now ready for application. Currently we are looking for other means to acquire sufficient work force to perform the originally planned assessment.
