Ensemble forecasts in reproducing kernel Hilbert space family

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Dufée, Benjamin, Hug, Bérenger, Mémin, Etienne, Tissot, Gilles
Format: Preprint
Published: 2022
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914625007648768
author Dufée, Benjamin
Hug, Bérenger
Mémin, Etienne
Tissot, Gilles
author_facet Dufée, Benjamin
Hug, Bérenger
Mémin, Etienne
Tissot, Gilles
contents A methodological framework for ensemble-based estimation and simulation of high dimensional dynamical systems such as the oceanic or atmospheric flows is proposed. To that end, the dynamical system is embedded in a family of reproducing kernel Hilbert spaces (RKHS) with kernel functions driven by the dynamics. In the RKHS family, the Koopman and Perron-Frobenius operators are unitary and uniformly continuous. This property warrants they can be expressed in exponential series of diagonalizable bounded evolution operators defined from their infinitesimal generators. Access to Lyapunov exponents and to exact ensemble based expressions of the tangent linear dynamics are directly available as well. The RKHS family enables us the devise of strikingly simple ensemble data assimilation methods for trajectory reconstructions in terms of constant-in-time linear combinations of trajectory samples. Such an embarrassingly simple strategy is made possible through a fully justified superposition principle ensuing from several fundamental theorems.
format Preprint
id arxiv_https___arxiv_org_abs_2207_14653
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle Ensemble forecasts in reproducing kernel Hilbert space family
Dufée, Benjamin
Hug, Bérenger
Mémin, Etienne
Tissot, Gilles
Mathematical Physics
Machine Learning
Dynamical Systems
Computational Physics
Data Analysis, Statistics and Probability
A methodological framework for ensemble-based estimation and simulation of high dimensional dynamical systems such as the oceanic or atmospheric flows is proposed. To that end, the dynamical system is embedded in a family of reproducing kernel Hilbert spaces (RKHS) with kernel functions driven by the dynamics. In the RKHS family, the Koopman and Perron-Frobenius operators are unitary and uniformly continuous. This property warrants they can be expressed in exponential series of diagonalizable bounded evolution operators defined from their infinitesimal generators. Access to Lyapunov exponents and to exact ensemble based expressions of the tangent linear dynamics are directly available as well. The RKHS family enables us the devise of strikingly simple ensemble data assimilation methods for trajectory reconstructions in terms of constant-in-time linear combinations of trajectory samples. Such an embarrassingly simple strategy is made possible through a fully justified superposition principle ensuing from several fundamental theorems.
title Ensemble forecasts in reproducing kernel Hilbert space family
topic Mathematical Physics
Machine Learning
Dynamical Systems
Computational Physics
Data Analysis, Statistics and Probability
url https://arxiv.org/abs/2207.14653