Models used for climate predictions, i.e. time ranging from sub-seasonal, seasonal, inter-annual, to decadal scale, tend to show a bias which evolves with lead time after initialization of the model simulation. The lead time dependent bias, i.e. drift, results from initial states differing from the model climatology.
The workshop aimed at the better understanding of physical processes of initial shock and drift, as well as reviewing current strategies of drift and bias adjustment in climate predictions. The goal was also to formulate recommendations for future research activities.
The workshop brought together researchers with a modeling or statistical background to jointly discuss the different perspectives of bias adjustment. Several MiKlip researchers have contributed to the workshop.
Find more details on the workshop website. Under the heading "Agenda" you find links to the talks and posters presented at the meeting.