Process-oriented models of autumn leaf phenology: ways to sound calibration and implications of uncertain projections

Meier, Michael; Bigler, Christof

doi:https://doi.org/10.5194/gmd-16-7171-2023

Articles | Volume 16, issue 23

https://doi.org/10.5194/gmd-16-7171-2023

© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/gmd-16-7171-2023

© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 16, issue 23

Model evaluation paper

|

11 Dec 2023

Model evaluation paper |

| 11 Dec 2023

Process-oriented models of autumn leaf phenology: ways to sound calibration and implications of uncertain projections

Michael Meier and Christof Bigler

Supplement

https://doi.org/10.5194/gmd-16-7171-2023-supplement

Data sets

Modelled past autumn leaf phenology of deciduous trees Michael Meier Christof Bigler https://doi.org/10.5061/dryad.dv41ns22k

Projected future autumn leaf phenology of deciduous trees Michael Meier Christof Bigler https://doi.org/10.5061/dryad.mw6m90613

Model code and software

Process-oriented models of autumn leaf phenology Michael Meier https://doi.org/10.5281/zenodo.7188160

Short summary

We analyzed >2.3 million calibrations and 39 million projections of leaf coloration models, considering 21 models, 5 optimization algorithms, ≥7 sampling procedures, and 26 climate scenarios. Models based on temperature, day length, and leaf unfolding performed best, especially when calibrated with generalized simulated annealing and systematically balanced or stratified samples. Projected leaf coloration shifts between −13 and +20 days by 2080–2099.