Skillful forecasts were generated using four model classes - process models, machine learning, empirical, and multi-model ensembles.

35-40% of models outperformed null models and the most successful models accounted for more sources of uncertainty (process, parameter, initial conditions, observational, drivers) with poor models often over-confident in their forecasts.

Are you interested in trying to forecast ecological variables?🔮
Or want to teach some key forecasting concepts?
Our #open-source, R-based, tutorial materials are available via GitHub, along with a short paper at #jose jose.theoj.org/papers/10.21...
@careylab.bsky.social @cboettig.bsky.social