As errors initiating both from imperfect initial conditions and numerical models amplify due to the chaotic nature of the atmosphere, weather predictability is limited. The limits of predictability were first explored climatologically, in a time mean sense. With the advent of ensemble forecasting where a set of integrations are carried out in place of a single forecast to represent plausible scenarios given initial condition and model related errors, the case dependent assessment of predictability, in the form of ensemble-based probabilistic forecasts, became a reality. This presentation will review concepts and methods developed for the key components of ensemble forecast systems such as the estimation and sampling of initial uncertainty for the generation of initial perturbations, and the representation of model related uncertainty.

This event is part of the eventgroup Meteorology Colloquium Karlsruhe

Homepage
https://www.imk.kit.edu/88.php

Speaker
Dr. Zoltan Toth

Global Systems Division, Earth Science Lab., Oceanic and Atm.Research
National Oceanic and Atmo.Admin. (GSD/ESRL/OAR/NOAA) - USA

Organizer
IMK-TRO
Institute of Meteorology and Climate Research
KIT
Wolfgang-Gaede-Str. 1
76131 Karlsruhe
Tel: 0721 608 43356
Mail: imk-tro ∂does-not-exist.kit edu
https://www.imk-tro.kit.edu

Targetgroup
Interested / Everyone