Convective Meteorology (Mesoscale Dynamics)

Investigating the Optimal Design of Convection-Allowing Ensembles from a Model-Error Perspective

Jeff Duda

School of Meteorology

16 October 2015, 3:00 PM

National Weather Center, Room 5600
120 David L. Boren Blvd.
University of Oklahoma
Norman, OK

Ensemble forecasting at convection-allowing resolutions is becoming increasingly common as technology continues to improve and research into best design practices continue. Forecast error from an NWP model can come from errors in the initial conditions, the lateral boundary conditions that force the model (if a limited-area domain is used), and the internal model processes including the dynamics and parameterized physics (referred to as model error). My PhD research has focused on various ways to account for the latter in the context of optimal ensemble design at the convective scale.

This talk will begin with a brief history and overview of the theory of ensemble prediction before I delve into the various projects I have worked on as part of my PhD research. I will cover a method of accounting for microphysics uncertainty (mixed microphysics vs. perturbed parameter microphysics) and a method of stochastic-error representation (stochastic kinetic energy backscatter). I will spend the final half of the talk discussing current research into accounting for land-surface model uncertainty.

Convective Meteorology (Mesoscale Dynamics) Seminar Series website