Combining analog method and ensemble data assimilation: application to the Lorenz-63 chaotic system

PIERRE TANDEO; PIERRE AILLLIOT; JUAN RUIZ; ALEXIS HANNART; BERTRAND CHAPRON; ANNE CUZOL; VALERIE MONBET; ROBERT EASTON; RONAN FABLET

Machine Learning and Data Mining Approaches to Climate Science

Springer

Año: 2015; p. 1 - 10

Abstract Nowadays, ocean and atmosphere sciences face a deluge of data from space, in situ monitoring as well as numerical simulations. The availability of these different data sources offer new opportunities, still largely underexploited, to im- prove the understanding, modeling and reconstruction of geophysical dynamics. The classical way to reconstruct the space-time variations of a geophysical sys- tem from observations relies on multiple runs of the known dynamical model. This classical framework may have severe limitations such as its computational cost, the lack of adequacy of the model and/or its parameterization with observed data, modeling uncertainties, etc... In this paper, we explore an alternative approach and develop a fully data-driven framework, which combines machine learning and statis- tical sampling to simulate the dynamics of complex system. As a proof concept, we address the assimilation of the chaotic Lorenz-63 model.We demonstrate that a nonparametric sampler from a catalog of historical datasets, namely a nearest neighbor or analog sampler, combined with a classical stochastic data assimilation scheme, the ensemble Kalman filter and smoother, reach state-of-the-art performances, without online evaluations of the