From Many to One: Consensus Inference in a MIP

Abstract A Model Intercomparison Project (MIP) consists of teams who estimate the same underlying quantity (e.g., temperature projections to the year 2070). A simple average of the ensemble of the teams' outputs gives a consensus estimate, but it does not recognize that some outputs are more va...

Full description

Saved in:
Bibliographic Details
Main Authors: Noel Cressie, Michael Bertolacci, Andrew Zammit‐Mangion
Format: Article
Language:English
Published: Wiley 2022-07-01
Series:Geophysical Research Letters
Subjects:
Online Access:https://doi.org/10.1029/2022GL098277
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849712527163785216
author Noel Cressie
Michael Bertolacci
Andrew Zammit‐Mangion
author_facet Noel Cressie
Michael Bertolacci
Andrew Zammit‐Mangion
author_sort Noel Cressie
collection DOAJ
description Abstract A Model Intercomparison Project (MIP) consists of teams who estimate the same underlying quantity (e.g., temperature projections to the year 2070). A simple average of the ensemble of the teams' outputs gives a consensus estimate, but it does not recognize that some outputs are more variable than others. Statistical analysis of variance (ANOVA) models offer a way to obtain a weighted frequentist consensus estimate of outputs with a variance that is the smallest possible. Modulo dependence between MIP outputs, the ANOVA approach weights a team's output inversely proportional to its variance, from which optimally weighted estimates follow. ANOVA weights can also provide a prior distribution for Bayesian Model Averaging of the MIP outputs when external evaluation data are available. We use a MIP of carbon‐dioxide‐flux inversions to illustrate the ANOVA‐based weighting and subsequent frequentist consensus inferences.
format Article
id doaj-art-edf2a7d76ee44f00bac93399587a027e
institution DOAJ
issn 0094-8276
1944-8007
language English
publishDate 2022-07-01
publisher Wiley
record_format Article
series Geophysical Research Letters
spelling doaj-art-edf2a7d76ee44f00bac93399587a027e2025-08-20T03:14:13ZengWileyGeophysical Research Letters0094-82761944-80072022-07-014914n/an/a10.1029/2022GL098277From Many to One: Consensus Inference in a MIPNoel Cressie0Michael Bertolacci1Andrew Zammit‐Mangion2School of Mathematics and Applied Statistics University of Wollongong Wollongong NSW AustraliaSchool of Mathematics and Applied Statistics University of Wollongong Wollongong NSW AustraliaSchool of Mathematics and Applied Statistics University of Wollongong Wollongong NSW AustraliaAbstract A Model Intercomparison Project (MIP) consists of teams who estimate the same underlying quantity (e.g., temperature projections to the year 2070). A simple average of the ensemble of the teams' outputs gives a consensus estimate, but it does not recognize that some outputs are more variable than others. Statistical analysis of variance (ANOVA) models offer a way to obtain a weighted frequentist consensus estimate of outputs with a variance that is the smallest possible. Modulo dependence between MIP outputs, the ANOVA approach weights a team's output inversely proportional to its variance, from which optimally weighted estimates follow. ANOVA weights can also provide a prior distribution for Bayesian Model Averaging of the MIP outputs when external evaluation data are available. We use a MIP of carbon‐dioxide‐flux inversions to illustrate the ANOVA‐based weighting and subsequent frequentist consensus inferences.https://doi.org/10.1029/2022GL098277analysis of variance (ANOVA)Model Intercomparison Project (MIP)multi model ensemblestatistically optimal weightsSUPE‐ANOVA frameworkuncertainty quantification
spellingShingle Noel Cressie
Michael Bertolacci
Andrew Zammit‐Mangion
From Many to One: Consensus Inference in a MIP
Geophysical Research Letters
analysis of variance (ANOVA)
Model Intercomparison Project (MIP)
multi model ensemble
statistically optimal weights
SUPE‐ANOVA framework
uncertainty quantification
title From Many to One: Consensus Inference in a MIP
title_full From Many to One: Consensus Inference in a MIP
title_fullStr From Many to One: Consensus Inference in a MIP
title_full_unstemmed From Many to One: Consensus Inference in a MIP
title_short From Many to One: Consensus Inference in a MIP
title_sort from many to one consensus inference in a mip
topic analysis of variance (ANOVA)
Model Intercomparison Project (MIP)
multi model ensemble
statistically optimal weights
SUPE‐ANOVA framework
uncertainty quantification
url https://doi.org/10.1029/2022GL098277
work_keys_str_mv AT noelcressie frommanytooneconsensusinferenceinamip
AT michaelbertolacci frommanytooneconsensusinferenceinamip
AT andrewzammitmangion frommanytooneconsensusinferenceinamip