Using global Bayesian optimization in ensemble data assimilation: parameter estimation, tuning localization and inflation, or all of the above

Global Bayesian optimization (GBO) is a derivative-free optimization method that is used widely in the tech-industry to optimize objective functions that are expensive to evaluate, numerically or otherwise. We discuss the use of GBO in ensemble data assimilation (DA), where the goal is to update the...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Spencer Lunderman, Matthias Morzfeld, Derek J. Posselt
Formato:	article
Lenguaje:	EN
Publicado:	Taylor & Francis Group 2021
Materias:	data assimilation global bayesian optimization parameter estimation ensemble kalman filtering Oceanography GC1-1581 Meteorology. Climatology QC851-999
Acceso en línea:	https://doaj.org/article/028de1b4d61b4d8c8fb124ca91f85a41
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:028de1b4d61b4d8c8fb124ca91f85a41
record_format	dspace
spelling	oai:doaj.org-article:028de1b4d61b4d8c8fb124ca91f85a412021-12-01T14:40:58ZUsing global Bayesian optimization in ensemble data assimilation: parameter estimation, tuning localization and inflation, or all of the above1600-087010.1080/16000870.2021.1924952https://doaj.org/article/028de1b4d61b4d8c8fb124ca91f85a412021-01-01T00:00:00Zhttp://dx.doi.org/10.1080/16000870.2021.1924952https://doaj.org/toc/1600-0870Global Bayesian optimization (GBO) is a derivative-free optimization method that is used widely in the tech-industry to optimize objective functions that are expensive to evaluate, numerically or otherwise. We discuss the use of GBO in ensemble data assimilation (DA), where the goal is to update the state of a numerical model in view of noisy observations. Specifically, we consider three tasks: (i) the estimation of model parameters; (ii) the tuning of localization and inflation in ensemble DA; (iii) doing both, i.e. estimating model parameters while simultaneously tuning the localization and inflation of the ensemble DA. For all three tasks, the GBO works ‘offline’, i.e. a set of ‘training’ observations are used within GBO to determine appropriate model or localization/inflation parameters, which are subsequently deployed within an ensemble DA system. Because of the offline nature of the technique, GBO can easily be combined with existing DA systems and it can effectively decouple (nearly) linear/Gaussian aspects of a problem from highly nonlinear/non-Gaussian ones. We illustrate the use of GBO in simple numerical experiments with the classical Lorenz problems. Our main goals are to introduce GBO in the context of ensemble DA and to spark an interest in GBO and its uses for streamlining important tasks in ensemble DA.Spencer LundermanMatthias MorzfeldDerek J. PosseltTaylor & Francis Grouparticledata assimilationglobal bayesian optimizationparameter estimationensemble kalman filteringOceanographyGC1-1581Meteorology. ClimatologyQC851-999ENTellus: Series A, Dynamic Meteorology and Oceanography, Vol 73, Iss 1, Pp 1-16 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	data assimilation global bayesian optimization parameter estimation ensemble kalman filtering Oceanography GC1-1581 Meteorology. Climatology QC851-999
spellingShingle	data assimilation global bayesian optimization parameter estimation ensemble kalman filtering Oceanography GC1-1581 Meteorology. Climatology QC851-999 Spencer Lunderman Matthias Morzfeld Derek J. Posselt Using global Bayesian optimization in ensemble data assimilation: parameter estimation, tuning localization and inflation, or all of the above
description	Global Bayesian optimization (GBO) is a derivative-free optimization method that is used widely in the tech-industry to optimize objective functions that are expensive to evaluate, numerically or otherwise. We discuss the use of GBO in ensemble data assimilation (DA), where the goal is to update the state of a numerical model in view of noisy observations. Specifically, we consider three tasks: (i) the estimation of model parameters; (ii) the tuning of localization and inflation in ensemble DA; (iii) doing both, i.e. estimating model parameters while simultaneously tuning the localization and inflation of the ensemble DA. For all three tasks, the GBO works ‘offline’, i.e. a set of ‘training’ observations are used within GBO to determine appropriate model or localization/inflation parameters, which are subsequently deployed within an ensemble DA system. Because of the offline nature of the technique, GBO can easily be combined with existing DA systems and it can effectively decouple (nearly) linear/Gaussian aspects of a problem from highly nonlinear/non-Gaussian ones. We illustrate the use of GBO in simple numerical experiments with the classical Lorenz problems. Our main goals are to introduce GBO in the context of ensemble DA and to spark an interest in GBO and its uses for streamlining important tasks in ensemble DA.
format	article
author	Spencer Lunderman Matthias Morzfeld Derek J. Posselt
author_facet	Spencer Lunderman Matthias Morzfeld Derek J. Posselt
author_sort	Spencer Lunderman
title	Using global Bayesian optimization in ensemble data assimilation: parameter estimation, tuning localization and inflation, or all of the above
title_short	Using global Bayesian optimization in ensemble data assimilation: parameter estimation, tuning localization and inflation, or all of the above
title_full	Using global Bayesian optimization in ensemble data assimilation: parameter estimation, tuning localization and inflation, or all of the above
title_fullStr	Using global Bayesian optimization in ensemble data assimilation: parameter estimation, tuning localization and inflation, or all of the above
title_full_unstemmed	Using global Bayesian optimization in ensemble data assimilation: parameter estimation, tuning localization and inflation, or all of the above
title_sort	using global bayesian optimization in ensemble data assimilation: parameter estimation, tuning localization and inflation, or all of the above
publisher	Taylor & Francis Group
publishDate	2021
url	https://doaj.org/article/028de1b4d61b4d8c8fb124ca91f85a41
work_keys_str_mv	AT spencerlunderman usingglobalbayesianoptimizationinensembledataassimilationparameterestimationtuninglocalizationandinflationoralloftheabove AT matthiasmorzfeld usingglobalbayesianoptimizationinensembledataassimilationparameterestimationtuninglocalizationandinflationoralloftheabove AT derekjposselt usingglobalbayesianoptimizationinensembledataassimilationparameterestimationtuninglocalizationandinflationoralloftheabove
_version_	1718405036428492800

Using global Bayesian optimization in ensemble data assimilation: parameter estimation, tuning localization and inflation, or all of the above

Ejemplares similares