"Synchronisation for data assimilation: an ensemble framework"

［ 2017年03月01日 ］

RIKEN International Symposium on Data Assimilation 2017
"Synchronisation for data assimilation: an ensemble framework"

Current data assimilation methods still face problems in strongly nonlinear cases. A promising solution is a particle filter, which provides a representation of the model probability density function by a discrete set of particles. However, the basic particle filter does not work in high-dimensional cases. The performance can be improved by considering the proposal density freedom. A potential choice of proposal density might come from the synchronisation theory, in which one tries to synchronise the model with the true evolution of a system via the observations. In practice, an extra term is added to the model equations that hampers growth of instabilities on the synchronisation manifold. In this work, an innovative methodology is proposed: an ensemble-based synchronisation scheme. Tests were performed using Lorenz96 model for 20, 100 and 1000-dimension systems. Results show synchronisation errors of an order of magnitude of 10(-2), suggesting that the scheme is a promising tool to steer model states to the truth. Potential benefits are expected for the numerical weather prediction, as the future goal is to test this methodology on each particle in a particle filter system, so obtaining a fully nonlinear data assimilation scheme to allow better forecasts.
http://www.data-assimilation.riken.jp/risda2017/

講師プロフィール
名前:Flavia Rodrigues Pinheiro	所属:University of Reading
略歴:外部ページ

eラーニングアーカイブ カテゴリー