I attended the ‘Workshop on Uncertainties’ organised at the Potsdam Institute for Climate Impact Research (PIK) Potsdam last week from 29 September to 2 October 2020. Keynote lectures from Sune Rasmussen, Maarten Blaauw, Nadine Berner, and Rebecca Morisson covered several interesting aspects of uncertainties. Topics included: how (and why) ice core uncertainties are set conservatively (Rasmussen), how we can use Bayesian principles to infer age models from measured radiometric data (Blaauw), structural aspects of change point detection (Berner), and learning unknown extensions of models to fit data better (Morisson).

The workshop was also a great chance to see young researchers in action! In particular, Keno Reichers (PIK) presented an creative way to determine the lag between the onset of abrupt transitions in two proxy records obtained from Greenland ice cores. Tobias Braun (PIK) presented a new way to obtain seasonal variability from stalagmite proxy records which have been modeled using COPRA. Annabel Wolf presented work from her newly published paper (link) which discusses how rainfall in central Vietnam is controlled by two moisture sources, viz. the Indian Ocean and the South China Sea, as revealed by stable oxygen isotope data.

Torben Kunz (AWI) presented recent work (link) which successfully decomposes the power spectrum of a paleoclimatic proxy into components that account for the different factors that go into estimating the proxy record from the measurements. The study does assume, however, that there is no age modeling involved but the results do look promising! The next extension of their approach, which would ideally include age modeling as well, is something to look forward to.

In a COVID-19 world, most of the participants acknowledged that it felt good to be meeting in persona, notwithstanding face masks, physical distancing, et al.