Effect of Data Assimilation Parameters on The Optimized Surface CO2 Flux in Asia

Hyunjung Kim, Hyun Mee Kim, Jinwoong Kim, Chun Ho Cho

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

In this study, CarbonTracker, an inverse modeling system based on the ensemble Kalman filter, was used to evaluate the effects of data assimilation parameters (assimilation window length and ensemble size) on the estimation of surface CO2 fluxes in Asia. Several experiments with different parameters were conducted, and the results were verified using CO2 concentration observations. The assimilation window lengths tested were 3, 5, 7, and 10 weeks, and the ensemble sizes were 100, 150, and 300. Therefore, a total of 12 experiments using combinations of these parameters were conducted. The experimental period was from January 2006 to December 2009. Differences between the optimized surface CO2 fluxes of the experiments were largest in the Eurasian Boreal (EB) area, followed by Eurasian Temperate (ET) and Tropical Asia (TA), and were larger in boreal summer than in boreal winter. The effect of ensemble size on the optimized biosphere flux is larger than the effect of the assimilation window length in Asia, but the importance of them varies in specific regions in Asia. The optimized biosphere flux was more sensitive to the assimilation window length in EB, whereas it was sensitive to the ensemble size as well as the assimilation window length in ET. The larger the ensemble size and the shorter the assimilation window length, the larger the uncertainty (i.e., spread of ensemble) of optimized surface CO2 fluxes. The 10-week assimilation window and 300 ensemble size were the optimal configuration for CarbonTracker in the Asian region based on several verifications using CO2 concentration measurements.

Original languageEnglish
JournalAsia-Pacific Journal of Atmospheric Sciences
Volume54
Issue number1
DOIs
Publication statusPublished - 2018 Feb 1

Fingerprint

data assimilation
biosphere
experiment
Kalman filter
Asia
assimilation
effect
parameter
winter
summer
modeling

All Science Journal Classification (ASJC) codes

  • Atmospheric Science

Cite this

@article{cb9ebb2554c3432ca8d727cc1a41af23,
title = "Effect of Data Assimilation Parameters on The Optimized Surface CO2 Flux in Asia",
abstract = "In this study, CarbonTracker, an inverse modeling system based on the ensemble Kalman filter, was used to evaluate the effects of data assimilation parameters (assimilation window length and ensemble size) on the estimation of surface CO2 fluxes in Asia. Several experiments with different parameters were conducted, and the results were verified using CO2 concentration observations. The assimilation window lengths tested were 3, 5, 7, and 10 weeks, and the ensemble sizes were 100, 150, and 300. Therefore, a total of 12 experiments using combinations of these parameters were conducted. The experimental period was from January 2006 to December 2009. Differences between the optimized surface CO2 fluxes of the experiments were largest in the Eurasian Boreal (EB) area, followed by Eurasian Temperate (ET) and Tropical Asia (TA), and were larger in boreal summer than in boreal winter. The effect of ensemble size on the optimized biosphere flux is larger than the effect of the assimilation window length in Asia, but the importance of them varies in specific regions in Asia. The optimized biosphere flux was more sensitive to the assimilation window length in EB, whereas it was sensitive to the ensemble size as well as the assimilation window length in ET. The larger the ensemble size and the shorter the assimilation window length, the larger the uncertainty (i.e., spread of ensemble) of optimized surface CO2 fluxes. The 10-week assimilation window and 300 ensemble size were the optimal configuration for CarbonTracker in the Asian region based on several verifications using CO2 concentration measurements.",
author = "Hyunjung Kim and Kim, {Hyun Mee} and Jinwoong Kim and Cho, {Chun Ho}",
year = "2018",
month = "2",
day = "1",
doi = "10.1007/s13143-017-0049-9",
language = "English",
volume = "54",
journal = "Asia-Pacific Journal of Atmospheric Sciences",
issn = "1976-7633",
publisher = "Korean Meteorological Society",
number = "1",

}

Effect of Data Assimilation Parameters on The Optimized Surface CO2 Flux in Asia. / Kim, Hyunjung; Kim, Hyun Mee; Kim, Jinwoong; Cho, Chun Ho.

In: Asia-Pacific Journal of Atmospheric Sciences, Vol. 54, No. 1, 01.02.2018.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Effect of Data Assimilation Parameters on The Optimized Surface CO2 Flux in Asia

AU - Kim, Hyunjung

AU - Kim, Hyun Mee

AU - Kim, Jinwoong

AU - Cho, Chun Ho

PY - 2018/2/1

Y1 - 2018/2/1

N2 - In this study, CarbonTracker, an inverse modeling system based on the ensemble Kalman filter, was used to evaluate the effects of data assimilation parameters (assimilation window length and ensemble size) on the estimation of surface CO2 fluxes in Asia. Several experiments with different parameters were conducted, and the results were verified using CO2 concentration observations. The assimilation window lengths tested were 3, 5, 7, and 10 weeks, and the ensemble sizes were 100, 150, and 300. Therefore, a total of 12 experiments using combinations of these parameters were conducted. The experimental period was from January 2006 to December 2009. Differences between the optimized surface CO2 fluxes of the experiments were largest in the Eurasian Boreal (EB) area, followed by Eurasian Temperate (ET) and Tropical Asia (TA), and were larger in boreal summer than in boreal winter. The effect of ensemble size on the optimized biosphere flux is larger than the effect of the assimilation window length in Asia, but the importance of them varies in specific regions in Asia. The optimized biosphere flux was more sensitive to the assimilation window length in EB, whereas it was sensitive to the ensemble size as well as the assimilation window length in ET. The larger the ensemble size and the shorter the assimilation window length, the larger the uncertainty (i.e., spread of ensemble) of optimized surface CO2 fluxes. The 10-week assimilation window and 300 ensemble size were the optimal configuration for CarbonTracker in the Asian region based on several verifications using CO2 concentration measurements.

AB - In this study, CarbonTracker, an inverse modeling system based on the ensemble Kalman filter, was used to evaluate the effects of data assimilation parameters (assimilation window length and ensemble size) on the estimation of surface CO2 fluxes in Asia. Several experiments with different parameters were conducted, and the results were verified using CO2 concentration observations. The assimilation window lengths tested were 3, 5, 7, and 10 weeks, and the ensemble sizes were 100, 150, and 300. Therefore, a total of 12 experiments using combinations of these parameters were conducted. The experimental period was from January 2006 to December 2009. Differences between the optimized surface CO2 fluxes of the experiments were largest in the Eurasian Boreal (EB) area, followed by Eurasian Temperate (ET) and Tropical Asia (TA), and were larger in boreal summer than in boreal winter. The effect of ensemble size on the optimized biosphere flux is larger than the effect of the assimilation window length in Asia, but the importance of them varies in specific regions in Asia. The optimized biosphere flux was more sensitive to the assimilation window length in EB, whereas it was sensitive to the ensemble size as well as the assimilation window length in ET. The larger the ensemble size and the shorter the assimilation window length, the larger the uncertainty (i.e., spread of ensemble) of optimized surface CO2 fluxes. The 10-week assimilation window and 300 ensemble size were the optimal configuration for CarbonTracker in the Asian region based on several verifications using CO2 concentration measurements.

UR - http://www.scopus.com/inward/record.url?scp=85030098625&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85030098625&partnerID=8YFLogxK

U2 - 10.1007/s13143-017-0049-9

DO - 10.1007/s13143-017-0049-9

M3 - Article

AN - SCOPUS:85030098625

VL - 54

JO - Asia-Pacific Journal of Atmospheric Sciences

JF - Asia-Pacific Journal of Atmospheric Sciences

SN - 1976-7633

IS - 1

ER -