Model validation should provide a quantification of the model uncertainties, a reliable review of the forecast results and the possibility for a direct comparison between results of different model simulations.
Atmospheric model validation is presently evolving more and more towards the development and comparison of so-called “process-oriented variables”: deficiencies in reproducing temperature and precipitation fields for example have been identified in terms of model restrictions in reproducing atmospheric circulation variability, leading in particular to substantial differences between models and observation in circulation-climate relationships. Representing the most important processes of atmospheric dynamics (large-scale circulation patterns, planetary and gravity waves) is therefore crucial for climate modelling especially in the range of ten years.
Our proposal (VADY – validation of atmospheric dynamics) to address this issue with respect to model validation includes various approaches focussing on atmospheric wave activity (planetary and gravity waves) in different atmospheric layers and on different spatial scales as well as on circulation types, dynamical modes and teleconnections in the atmosphere. Multivariate statistics as well as recently developed algorithms will be used for the determination and comparison of some few variables mostly easy to calculate in order to allow a relatively quick answer to questions about the representation of atmospheric dynamics in the model.
Comparison of MiKlip-model runs with re-analysis data & historical satellite data concerning the following parameters:
University of Augsburg
Comparison of MiKlip-model runs with re-analysis data concerning the following objectives based on multi-variate statistics:
Atmospheric circulation types