Waves to Weather

Breadcrumb Navigation


A1 - Upscale impact of diabatic processes from convective to near-hemispheric scale

Principal Investigators: Prof. Dr. George C. Craig, Prof. Dr. Volkmar Wirth, Dr. Michael Riemer

Other researchers: Dr. Tobias Selz (W2W member), Lotte Bierdel (Postdoc), Marlene Baumgart (PhD), Paolo Ghinassi (PhD), Anne Martin (Master)

In this project we will investigate the process of upscale propagation of uncertainty in the atmosphere over three orders of magnitude in spatial scale, from convective clouds to hemispheric waves.

This will be possible by combining expertise of three partners at two universities and by taking advantage of recent developments in numerical atmospheric modeling (ICON) and stochastic parameterization (Plant-Craig). The non-hydrostatic ICON model allows for local grid refinements while the Plant-Craig convection scheme is able to emulate convective uncertainty at non-convective permitting resolutions. These two tools will form the basis for a series of error growth experiments to address open questions about basic characteristics, mechanisms and the practical importance of upscale error growth in mid-range global weather prediction.

Diagnostic tools already successfully applied in uncertainty propagation studies will be further developed and jointly applied to the considered cases to attribute physical processes to error growth and improve our understanding of sources and limitations of atmospheric predictability.

A novel diagnostic will be developed which allows one to diagnose wave activity for finite amplitude Rossby wave packets; it will be applied in order to quantify forecast errors and ensemble spread on the large-scale end of the spectrum.

This project will also contribute to a better conceptual understanding of the relation between two seemingly distinct perspectives on balanced dynamics, namely potential vorticity thinking and the wave activity framework. We will cooperate with visualization projects to facilitate the handling of five-dimensional ensemble data and extracting the relevant information.