Experiences fom the application of a parameter estimation and identifiability analysis methodology to the operational street pollution model (OSPM)

Thor Bjørn Ottosen, Matthias Ketzel, Ole Hertel, Henrik Skov, Jørgen Brandt, Ruwim Berkowicz, Konstantinos Kakosimos

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Uncertainty and sensitivity analysis can potentially increase the transparency in the modelling process and guide research in the relationship between model and data. The uncertainty and sensitivity of the Operational Street Pollution Model (OSPM), being an example of a semi-parameterised air quality model, have not been studied before, and it is therefore the aim to explore the potential advantages of this type of analyses on atmospheric models. An iterative parameter estimation and identifiability analysis methodology along with two different data splitting methodologies were chosen for the present study. The results show that this type of methodology can be informative applied to an atmospheric model, in that the methodology successfully balances the model-measurement errors among the different streets and the different species. Moreover, the results indicate where future research effort in model improvement should be directed, with respect to parameterisations and model parameter uncertainty.

Original languageEnglish
Title of host publicationHARMO 2014 - 16th International Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, Proceedings
PublisherConference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes
Pages454-458
Number of pages5
Publication statusPublished - 2014
Event16th International Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, HARMO 2014 - Varna, Bulgaria
Duration: 8 Sep 201411 Sep 2014

Other

Other16th International Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, HARMO 2014
CountryBulgaria
CityVarna
Period8/9/1411/9/14

Fingerprint

Identifiability
Pollution
Parameter estimation
Parameter Estimation
pollution
methodology
Methodology
Model
Measurement Error Model
Uncertainty Analysis
Air Quality
uncertainty analysis
Model Uncertainty
Parameter Uncertainty
Process Modeling
Transparency
transparency
Parameterization
Uncertainty analysis
Sensitivity Analysis

Keywords

  • Data splitting
  • Exploratory data analysis
  • OSPM
  • Sensitivity analysis
  • Uncertainty analysis

ASJC Scopus subject areas

  • Atmospheric Science
  • Pollution
  • Modelling and Simulation

Cite this

Ottosen, T. B., Ketzel, M., Hertel, O., Skov, H., Brandt, J., Berkowicz, R., & Kakosimos, K. (2014). Experiences fom the application of a parameter estimation and identifiability analysis methodology to the operational street pollution model (OSPM). In HARMO 2014 - 16th International Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, Proceedings (pp. 454-458). Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes.

Experiences fom the application of a parameter estimation and identifiability analysis methodology to the operational street pollution model (OSPM). / Ottosen, Thor Bjørn; Ketzel, Matthias; Hertel, Ole; Skov, Henrik; Brandt, Jørgen; Berkowicz, Ruwim; Kakosimos, Konstantinos.

HARMO 2014 - 16th International Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, Proceedings. Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, 2014. p. 454-458.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ottosen, TB, Ketzel, M, Hertel, O, Skov, H, Brandt, J, Berkowicz, R & Kakosimos, K 2014, Experiences fom the application of a parameter estimation and identifiability analysis methodology to the operational street pollution model (OSPM). in HARMO 2014 - 16th International Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, Proceedings. Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, pp. 454-458, 16th International Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, HARMO 2014, Varna, Bulgaria, 8/9/14.
Ottosen TB, Ketzel M, Hertel O, Skov H, Brandt J, Berkowicz R et al. Experiences fom the application of a parameter estimation and identifiability analysis methodology to the operational street pollution model (OSPM). In HARMO 2014 - 16th International Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, Proceedings. Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes. 2014. p. 454-458
Ottosen, Thor Bjørn ; Ketzel, Matthias ; Hertel, Ole ; Skov, Henrik ; Brandt, Jørgen ; Berkowicz, Ruwim ; Kakosimos, Konstantinos. / Experiences fom the application of a parameter estimation and identifiability analysis methodology to the operational street pollution model (OSPM). HARMO 2014 - 16th International Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, Proceedings. Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, 2014. pp. 454-458
@inproceedings{4636afb86a504ceb88089ae330241480,
title = "Experiences fom the application of a parameter estimation and identifiability analysis methodology to the operational street pollution model (OSPM)",
abstract = "Uncertainty and sensitivity analysis can potentially increase the transparency in the modelling process and guide research in the relationship between model and data. The uncertainty and sensitivity of the Operational Street Pollution Model (OSPM), being an example of a semi-parameterised air quality model, have not been studied before, and it is therefore the aim to explore the potential advantages of this type of analyses on atmospheric models. An iterative parameter estimation and identifiability analysis methodology along with two different data splitting methodologies were chosen for the present study. The results show that this type of methodology can be informative applied to an atmospheric model, in that the methodology successfully balances the model-measurement errors among the different streets and the different species. Moreover, the results indicate where future research effort in model improvement should be directed, with respect to parameterisations and model parameter uncertainty.",
keywords = "Data splitting, Exploratory data analysis, OSPM, Sensitivity analysis, Uncertainty analysis",
author = "Ottosen, {Thor Bj{\o}rn} and Matthias Ketzel and Ole Hertel and Henrik Skov and J{\o}rgen Brandt and Ruwim Berkowicz and Konstantinos Kakosimos",
year = "2014",
language = "English",
pages = "454--458",
booktitle = "HARMO 2014 - 16th International Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, Proceedings",
publisher = "Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes",

}

TY - GEN

T1 - Experiences fom the application of a parameter estimation and identifiability analysis methodology to the operational street pollution model (OSPM)

AU - Ottosen, Thor Bjørn

AU - Ketzel, Matthias

AU - Hertel, Ole

AU - Skov, Henrik

AU - Brandt, Jørgen

AU - Berkowicz, Ruwim

AU - Kakosimos, Konstantinos

PY - 2014

Y1 - 2014

N2 - Uncertainty and sensitivity analysis can potentially increase the transparency in the modelling process and guide research in the relationship between model and data. The uncertainty and sensitivity of the Operational Street Pollution Model (OSPM), being an example of a semi-parameterised air quality model, have not been studied before, and it is therefore the aim to explore the potential advantages of this type of analyses on atmospheric models. An iterative parameter estimation and identifiability analysis methodology along with two different data splitting methodologies were chosen for the present study. The results show that this type of methodology can be informative applied to an atmospheric model, in that the methodology successfully balances the model-measurement errors among the different streets and the different species. Moreover, the results indicate where future research effort in model improvement should be directed, with respect to parameterisations and model parameter uncertainty.

AB - Uncertainty and sensitivity analysis can potentially increase the transparency in the modelling process and guide research in the relationship between model and data. The uncertainty and sensitivity of the Operational Street Pollution Model (OSPM), being an example of a semi-parameterised air quality model, have not been studied before, and it is therefore the aim to explore the potential advantages of this type of analyses on atmospheric models. An iterative parameter estimation and identifiability analysis methodology along with two different data splitting methodologies were chosen for the present study. The results show that this type of methodology can be informative applied to an atmospheric model, in that the methodology successfully balances the model-measurement errors among the different streets and the different species. Moreover, the results indicate where future research effort in model improvement should be directed, with respect to parameterisations and model parameter uncertainty.

KW - Data splitting

KW - Exploratory data analysis

KW - OSPM

KW - Sensitivity analysis

KW - Uncertainty analysis

UR - http://www.scopus.com/inward/record.url?scp=84983197026&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84983197026&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84983197026

SP - 454

EP - 458

BT - HARMO 2014 - 16th International Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, Proceedings

PB - Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes

ER -