Maximum likelihood estimation of covariance parameters for Bayesian atmospheric trace gas surface flux inversions

Anna M. Michalak, Adam Hirsch, Lori Bruhwiler, Kevin Gurney, Wouter Peters, Pieter P. Tans

Research output: Contribution to journalArticle

77 Citations (Scopus)

Abstract

This paper introduces a Maximum Likelihood (ML) approach for estimating the statistical parameters required for the covariance matrices used in the solution of Bayesian inverse problems aimed at estimating surface fluxes of atmospheric trace gases. The method offers an objective methodology for populating the covariance matrices required in Bayesian inversions, thereby resulting in better estimates of the uncertainty associated with derived fluxes and minimizing the risk of inversions being biased by unrealistic covariance parameters. In addition, a method is presented for estimating the uncertainty associated with these covariance parameters. The ML method is demonstrated using a typical inversion setup with 22 flux regions and 75 observation stations from the National Oceanic and Atmospheric Administration-Climate Monitoring and Diagnostics Laboratory (NOAA-CMDL) Cooperative Air Sampling Network with available monthly averaged carbon dioxide data. Flux regions and observation locations are binned according to various characteristics, and the variances of the model-data mismatch and of the errors associated with the a priori flux distribution are estimated from the available data.

Original languageEnglish (US)
Article numberD24107
Pages (from-to)1-16
Number of pages16
JournalJournal of Geophysical Research: Atmospheres
Volume110
Issue number24
DOIs
StatePublished - Dec 27 2005
Externally publishedYes

Fingerprint

atmospheric gas
Maximum likelihood estimation
surface flux
trace gas
Gases
inversions
Fluxes
gases
estimating
Covariance matrix
air sampling
matrix
Maximum likelihood
inverse problem
carbon dioxide
Inverse problems
methodology
Carbon Dioxide
climate
monitoring

ASJC Scopus subject areas

  • Geochemistry and Petrology
  • Geophysics
  • Earth and Planetary Sciences (miscellaneous)
  • Space and Planetary Science
  • Atmospheric Science
  • Astronomy and Astrophysics
  • Oceanography

Cite this

Maximum likelihood estimation of covariance parameters for Bayesian atmospheric trace gas surface flux inversions. / Michalak, Anna M.; Hirsch, Adam; Bruhwiler, Lori; Gurney, Kevin; Peters, Wouter; Tans, Pieter P.

In: Journal of Geophysical Research: Atmospheres, Vol. 110, No. 24, D24107, 27.12.2005, p. 1-16.

Research output: Contribution to journalArticle

Michalak, Anna M. ; Hirsch, Adam ; Bruhwiler, Lori ; Gurney, Kevin ; Peters, Wouter ; Tans, Pieter P. / Maximum likelihood estimation of covariance parameters for Bayesian atmospheric trace gas surface flux inversions. In: Journal of Geophysical Research: Atmospheres. 2005 ; Vol. 110, No. 24. pp. 1-16.
@article{7fc380ca2eab441f83c26088d0af5be9,
title = "Maximum likelihood estimation of covariance parameters for Bayesian atmospheric trace gas surface flux inversions",
abstract = "This paper introduces a Maximum Likelihood (ML) approach for estimating the statistical parameters required for the covariance matrices used in the solution of Bayesian inverse problems aimed at estimating surface fluxes of atmospheric trace gases. The method offers an objective methodology for populating the covariance matrices required in Bayesian inversions, thereby resulting in better estimates of the uncertainty associated with derived fluxes and minimizing the risk of inversions being biased by unrealistic covariance parameters. In addition, a method is presented for estimating the uncertainty associated with these covariance parameters. The ML method is demonstrated using a typical inversion setup with 22 flux regions and 75 observation stations from the National Oceanic and Atmospheric Administration-Climate Monitoring and Diagnostics Laboratory (NOAA-CMDL) Cooperative Air Sampling Network with available monthly averaged carbon dioxide data. Flux regions and observation locations are binned according to various characteristics, and the variances of the model-data mismatch and of the errors associated with the a priori flux distribution are estimated from the available data.",
author = "Michalak, {Anna M.} and Adam Hirsch and Lori Bruhwiler and Kevin Gurney and Wouter Peters and Tans, {Pieter P.}",
year = "2005",
month = "12",
day = "27",
doi = "10.1029/2005JD005970",
language = "English (US)",
volume = "110",
pages = "1--16",
journal = "Journal of Geophysical Research: Atmospheres",
issn = "2169-897X",
publisher = "Wiley-Blackwell",
number = "24",

}

TY - JOUR

T1 - Maximum likelihood estimation of covariance parameters for Bayesian atmospheric trace gas surface flux inversions

AU - Michalak, Anna M.

AU - Hirsch, Adam

AU - Bruhwiler, Lori

AU - Gurney, Kevin

AU - Peters, Wouter

AU - Tans, Pieter P.

PY - 2005/12/27

Y1 - 2005/12/27

N2 - This paper introduces a Maximum Likelihood (ML) approach for estimating the statistical parameters required for the covariance matrices used in the solution of Bayesian inverse problems aimed at estimating surface fluxes of atmospheric trace gases. The method offers an objective methodology for populating the covariance matrices required in Bayesian inversions, thereby resulting in better estimates of the uncertainty associated with derived fluxes and minimizing the risk of inversions being biased by unrealistic covariance parameters. In addition, a method is presented for estimating the uncertainty associated with these covariance parameters. The ML method is demonstrated using a typical inversion setup with 22 flux regions and 75 observation stations from the National Oceanic and Atmospheric Administration-Climate Monitoring and Diagnostics Laboratory (NOAA-CMDL) Cooperative Air Sampling Network with available monthly averaged carbon dioxide data. Flux regions and observation locations are binned according to various characteristics, and the variances of the model-data mismatch and of the errors associated with the a priori flux distribution are estimated from the available data.

AB - This paper introduces a Maximum Likelihood (ML) approach for estimating the statistical parameters required for the covariance matrices used in the solution of Bayesian inverse problems aimed at estimating surface fluxes of atmospheric trace gases. The method offers an objective methodology for populating the covariance matrices required in Bayesian inversions, thereby resulting in better estimates of the uncertainty associated with derived fluxes and minimizing the risk of inversions being biased by unrealistic covariance parameters. In addition, a method is presented for estimating the uncertainty associated with these covariance parameters. The ML method is demonstrated using a typical inversion setup with 22 flux regions and 75 observation stations from the National Oceanic and Atmospheric Administration-Climate Monitoring and Diagnostics Laboratory (NOAA-CMDL) Cooperative Air Sampling Network with available monthly averaged carbon dioxide data. Flux regions and observation locations are binned according to various characteristics, and the variances of the model-data mismatch and of the errors associated with the a priori flux distribution are estimated from the available data.

UR - http://www.scopus.com/inward/record.url?scp=31644449372&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=31644449372&partnerID=8YFLogxK

U2 - 10.1029/2005JD005970

DO - 10.1029/2005JD005970

M3 - Article

VL - 110

SP - 1

EP - 16

JO - Journal of Geophysical Research: Atmospheres

JF - Journal of Geophysical Research: Atmospheres

SN - 2169-897X

IS - 24

M1 - D24107

ER -