2020
DOI: 10.1175/jcli-d-19-0769.1
|View full text |Cite
|
Sign up to set email alerts
|

A Bayesian Approach to Regional Decadal Predictability: Sparse Parameter Estimation in High-Dimensional Linear Inverse Models of High-Latitude Sea Surface Temperature Variability

Abstract: Stochastic reduced models are an important tool in climate systems whose many spatial and temporal scales cannot be fully discretized or underlying physics may not be fully accounted for. One form of reduced model, the linear inverse model (LIM), has been widely used for regional climate predictability studies - typically focusing more on tropical or mid-latitude studies. However, most LIM fitting techniques rely on point estimation techniques deriving from fluctuation-dissipation theory. In this methodologica… Show more

Help me understand this report
View preprint versions

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1

Citation Types

0
4
0

Year Published

2020
2020
2024
2024

Publication Types

Select...
4
1

Relationship

0
5

Authors

Journals

citations
Cited by 5 publications
(4 citation statements)
references
References 76 publications
(83 reference statements)
0
4
0
Order By: Relevance
“…52,53 An application in the field of oceanography, with regard to the parameter estimations of a high-dimensional linear stochastic differential equation (SDE), a model for high-latitude sea surface temperature variability, can be found in Ref. 54. There, Bayes' approach is used to provide optimal probabilistic estimates for a very highdimensional system, where the conditional PDF, typically called the conditional marginal likelihood, is assumed to be Gaussian.…”
Section: Articlementioning
confidence: 99%
“…52,53 An application in the field of oceanography, with regard to the parameter estimations of a high-dimensional linear stochastic differential equation (SDE), a model for high-latitude sea surface temperature variability, can be found in Ref. 54. There, Bayes' approach is used to provide optimal probabilistic estimates for a very highdimensional system, where the conditional PDF, typically called the conditional marginal likelihood, is assumed to be Gaussian.…”
Section: Articlementioning
confidence: 99%
“…Here we developed data-driven statistical emulators of the true eddy field for feeding them into the low-resolution model instead of the original high-resolution eddy fields. The number of statistical emulation methods has recently surged, including stochastic approaches in climate science (Penland and Matrosova, 2001;Strounine et al, 2010;Franzke et al, 2015;Chen et al, 2016;Palmer, 2019;Seleznev et al, 2019;Foster et al, 2020), as well as other machine-learning (deep learning) methods developed for fluid dynamics applications (Brunton et al, 2020;Bolton and Zanna, 2019). The detailed analysis of emulated eddy fields is beyond the scope of this study, and in the context of assessing the skill of our emulators we focus solely on one of the central problems in climate ocean model simulations, namely, the correct rectification of the eddy field's impact on the large-scale circulation.…”
Section: Statistical Emulation Of the Eddy Fieldmentioning
confidence: 99%
“…A widely‐used empirically based emulator is a linear inverse model (LIM; e.g., Penland & Sardeshmukh, 1995), which is attractive due to its computational efficiency, distinct timescale separation, and flexibility of calibration. Previous studies have found skillful decadal forecasts of regional sea surface temperatures using LIMs (e.g., Foster et al., 2020; Hawkins & Sutton, 2009). The LIM has also proven to be a suitable benchmark on decadal timescales that exceeds persistence and has comparable skill to Phase 5 of the Coupled Model Intercomparison Project model hindcasts for annual global surface temperature forecasts (Newman, 2013).…”
Section: Introductionmentioning
confidence: 96%