Atmospheric flows cover a scale range from ~ 0.1 mm (cloud droplets) to ~ 10 000 km (planet), and a comparable range of time scales. Computational models can only resolve part of this spectrum and their numerical discretizations modify the dynamics of scale interactions through associated truncation errors. In addition, any systematic impact of the unresolved scales on the resolved ones needs to be modeled by parametrization schemes, which introduce further model errors. At the same time, our insight into the current state of the atmosphere is limited by the spatio-temporal distribution of meteorological observations. In order to cope with the resulting uncertainties arising from discretization and modelling errors as well as from underdetermined initial and boundary conditions, data assimilation (DA) enables controlled adjustments of model-based forward simulations using incoming observational data in a judicious fashion.
DA combines such simulations with statistical views of observations and model output to minimize the model-to-data distance in a suitable norm. DA algorithms need to make explicit use of the underlying multi-scale nature of atmospheric .ows in order to be applicable in the presence of limited data sets and poor statistical resolutions. Currently this is achieved through adhoc techniques such as localization and ensemble in.ation, for which a satisfactory mathematical framework is as yet missing. At the same time, DA procedures also modify a model’s scale-dependent dynamics and may, e.g., affect the model’s balancing behavior with respect to fast internal gravity and acoustic wave modes – a problem with extensive history in numerical weather prediction. How can we quantify and correct for misrepresentations of scale interactions arising from underresolution, numerical truncation, and DA procedures?
Speaking in meteorological terms, this project aims at advanced data assimilation methods by synergetically connecting techniques of scale analysis, computational .uid dynamics, and advanced data .ltering. Methodologically speaking, we address the predictive modelling of the large scales of a system whose root model is known but inaccessible to direct computation due to a cascade of unresolvable spatio-temporal scales. We aim to exploit observational data, asymptotic characterizations of both the root model and the DA procedures, and asymptotically balanced numerical methods to control the resulting uncertainties.
Hittmeir, S. and Klein, R. and Li, J. and Titi, E. (2017) GLOBAL WELL-POSEDNESS FOR PASSIVELY TRANSPORTED NONLINEAR MOISTURE DYNAMICS WITH PHASE CHANGES. Nonlinearity, 30 (10). pp. 3676-3718.
Müller, A. and Névir, P. and Klein, R. (2017) Scale Dependent Analytical Investigation of the Dynamic State Index Concerning the Quasi-Geostrophic Theory. SFB 1114 Preprint . pp. 1-38. (Unpublished)
Acevedo, W. and de Wiljes, J. and Reich, S. (2017) Second-order accurate ensemble transform particle filters. SFB 1114 Preprint . (Submitted)
O'Kane, T.J. and Monselesan, D.P. and Risbey, J.S. and Horenko, I. and Franzke, Ch.L.E. (2017) On memory, dimension, and atmospheric teleconnection patterns. Math. Clim. Weather Forecast, 3 (1). pp. 1-27.
Horenko, I. and Gerber, S. and O'Kane, T.J. and Risbey, J.S. and Monselesan, D.P. (2017) On inference and validation of causality relations in climate teleconnections. In: Nonlinear and Stochastic Climate Dynamics. Cambridge University Press, pp. 184-208. ISBN 9781107118140
Chustagulprom, N. and Reich, S. and Reinhardt, M. (2016) A HYBRID ENSEMBLE TRANSFORM FILTER FOR HIGH DIMENSIONAL DYNAMICAL SYSTEMS. SIAM/ASA Journal on Uncertainty Quantification, 4 (1). pp. 592-608. ISSN 2166-2525
Klein, R. and Benacchio, T. (2016) A doubly blended model for multiscale atmospheric dynamics. Journal of the Atmospheric Sciences, 73 . pp. 1179-1186. ISSN Online: 1520-0469 Print: 0022-4928
Feireisl, E. and Klein, R. and Novotný, A. and Zatorska, E. (2016) On singular limits arising in the scale analysis of stratified fluid flows. Mathematical Models and Methods in Applied Sciences, World Scientific, 26 (3). pp. 419-443. ISSN Print: 0218-2025 Online: 1793-6314
Horenko, I. and Gerber, S. (2015) Improving clustering by imposing network information. Science Advances, 1 (7). ISSN 2375-2548