DS Talk with Han van der Aa, Professor in Process-Oriented Information System, from the Research Group Workflow Systems and Technology at the Faculty of Computer Science

(University of Cologne)

"Parametric Kalman Filtering: complex flow-dependent forecast uncertainties without ensembles"

This talk introduces a novel approach to estimate complex case-dependent uncertainties in high-dimensional atmospheric systems.

Instead of sampling forecast uncertainties from an ensemble, this parametric Kalman Filter (PKF) explicitly propagates key parameters of the uncertainty distribution. This enables (1) an understanding of forecast error dynamics and (2) improved data assimilation of sparse observations at low computational costs. After a general introduction to the method, a case-study application to Canadian wildfire smoke dispersal in the operational meteorological-air quality model GEM-MACH is presented.

Aus Böhmen nach Peru, aus Bremen nach Bordeaux

Netzwerke zwischen deutschen Protoindustrien, afrikanischen

Sklavenmärkten und Plantagenökonomien in den Amerikas

(c. 1680-1830)