Comparison of model estimates from an intra-city land use regression model with a national satellite-LUR and a regional Bayesian Maximum Entropy model, in estimating NO 2 for a birth cohort in Sydney, Australia

Christine T. Cowie, Frances Garden, Edward Jegasothy, Luke D. Knibbs, Ivan Hanigan, David Morley, Anna Hansell, Gerard Hoek, Guy B. Marks

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Background: Methods for estimating air pollutant exposures for epidemiological studies are becoming more complex in an effort to minimise exposure error and its associated bias. While land use regression (LUR) modelling is now an established method, there has been little comparison between LUR and other recent, more complex estimation methods. Our aim was to develop a LUR model to estimate intra-city exposures to nitrogen dioxide (NO 2 ) for a Sydney cohort, and to compare those with estimates from a national satellite-based LUR model (Sat-LUR) and a regional Bayesian Maximum Entropy (BME) model. Methods: Satellite-based LUR and BME estimates were obtained using existing models. We used methods consistent with the European Study of Cohorts for Air Pollution Effects (ESCAPE) methodology to develop LUR models for NO 2 and NOx. We deployed 46 Ogawa passive samplers across western Sydney during 2013/2014 and acquired data on land use, population density, and traffic volumes for the study area. Annual average NO 2 concentrations for 2013 were estimated for 947 addresses in the study area using the three models: standard LUR, Sat-LUR and a BME model. Agreement between the estimates from the three models was assessed using interclass correlation coefficient (ICC), Bland-Altman methods and correlation analysis (CC). Results: The NO 2 LUR model predicted 84% of spatial variability in annual mean NO 2 (RMSE: 1.2 ppb; cross-validated R 2 : 0.82) with predictors of major roads, population and dwelling density, heavy traffic and commercial land use. A separate model was developed that captured 92% of variability in NOx (RMSE 2.3 ppb; cross-validated R 2 : 0.90). The annual average NO 2 concentrations were 7.31 ppb (SD: 1.91), 7.01 ppb (SD: 1.92) and 7.90 ppb (SD: 1.85), for the LUR, Sat-LUR and BME models respectively. Comparing the standard LUR with Sat-LUR NO 2 cohort estimates, the mean estimates from the LUR were 4% higher than the Sat-LUR estimates, and the ICC was 0.73. The Pearson's correlation coefficients (CC) for the LUR vs Sat-LUR values were r = 0.73 (log-transformed data) and r = 0.69 (untransformed data). Comparison of the NO 2 cohort estimates from the LUR model with the BME blended model indicated that the LUR mean estimates were 8% lower than the BME estimates. The ICC for the LUR vs BME estimates was 0.73. The CC for the logged LUR vs BME estimates was r = 0.73 and for the unlogged estimates was r = 0.69. Conclusions: Our LUR models explained a high degree of spatial variability in annual mean NO 2 and NOx in western Sydney. The results indicate very good agreement between the intra-city LUR, national-scale sat-LUR, and regional BME models for estimating NO 2 for a cohort of children residing in Sydney, despite the different data inputs and differences in spatial scales of the models, providing confidence in their use in epidemiological studies.

LanguageEnglish
Pages24-34
Number of pages11
JournalEnvironmental Research
Volume174
DOIs
Publication statusPublished - Jul 2019

Fingerprint

Entropy
Land use
entropy
Parturition
Satellites
land use
Population Density
Epidemiologic Studies
comparison
city
Nitrogen Dioxide
Air Pollutants
Air Pollution
Cohort Studies
pollution effect

Cite this

Cowie, Christine T. ; Garden, Frances ; Jegasothy, Edward ; Knibbs, Luke D. ; Hanigan, Ivan ; Morley, David ; Hansell, Anna ; Hoek, Gerard ; Marks, Guy B. / Comparison of model estimates from an intra-city land use regression model with a national satellite-LUR and a regional Bayesian Maximum Entropy model, in estimating NO 2 for a birth cohort in Sydney, Australia. In: Environmental Research. 2019 ; Vol. 174. pp. 24-34.
@article{c4d752ee4e704a8c8c49c6bb269ddcad,
title = "Comparison of model estimates from an intra-city land use regression model with a national satellite-LUR and a regional Bayesian Maximum Entropy model, in estimating NO 2 for a birth cohort in Sydney, Australia",
abstract = "Background: Methods for estimating air pollutant exposures for epidemiological studies are becoming more complex in an effort to minimise exposure error and its associated bias. While land use regression (LUR) modelling is now an established method, there has been little comparison between LUR and other recent, more complex estimation methods. Our aim was to develop a LUR model to estimate intra-city exposures to nitrogen dioxide (NO 2 ) for a Sydney cohort, and to compare those with estimates from a national satellite-based LUR model (Sat-LUR) and a regional Bayesian Maximum Entropy (BME) model. Methods: Satellite-based LUR and BME estimates were obtained using existing models. We used methods consistent with the European Study of Cohorts for Air Pollution Effects (ESCAPE) methodology to develop LUR models for NO 2 and NOx. We deployed 46 Ogawa passive samplers across western Sydney during 2013/2014 and acquired data on land use, population density, and traffic volumes for the study area. Annual average NO 2 concentrations for 2013 were estimated for 947 addresses in the study area using the three models: standard LUR, Sat-LUR and a BME model. Agreement between the estimates from the three models was assessed using interclass correlation coefficient (ICC), Bland-Altman methods and correlation analysis (CC). Results: The NO 2 LUR model predicted 84{\%} of spatial variability in annual mean NO 2 (RMSE: 1.2 ppb; cross-validated R 2 : 0.82) with predictors of major roads, population and dwelling density, heavy traffic and commercial land use. A separate model was developed that captured 92{\%} of variability in NOx (RMSE 2.3 ppb; cross-validated R 2 : 0.90). The annual average NO 2 concentrations were 7.31 ppb (SD: 1.91), 7.01 ppb (SD: 1.92) and 7.90 ppb (SD: 1.85), for the LUR, Sat-LUR and BME models respectively. Comparing the standard LUR with Sat-LUR NO 2 cohort estimates, the mean estimates from the LUR were 4{\%} higher than the Sat-LUR estimates, and the ICC was 0.73. The Pearson's correlation coefficients (CC) for the LUR vs Sat-LUR values were r = 0.73 (log-transformed data) and r = 0.69 (untransformed data). Comparison of the NO 2 cohort estimates from the LUR model with the BME blended model indicated that the LUR mean estimates were 8{\%} lower than the BME estimates. The ICC for the LUR vs BME estimates was 0.73. The CC for the logged LUR vs BME estimates was r = 0.73 and for the unlogged estimates was r = 0.69. Conclusions: Our LUR models explained a high degree of spatial variability in annual mean NO 2 and NOx in western Sydney. The results indicate very good agreement between the intra-city LUR, national-scale sat-LUR, and regional BME models for estimating NO 2 for a cohort of children residing in Sydney, despite the different data inputs and differences in spatial scales of the models, providing confidence in their use in epidemiological studies.",
keywords = "Air pollution, Bayesian, Comparison, Exposure assessment, Land use regression, Satellite",
author = "Cowie, {Christine T.} and Frances Garden and Edward Jegasothy and Knibbs, {Luke D.} and Ivan Hanigan and David Morley and Anna Hansell and Gerard Hoek and Marks, {Guy B.}",
year = "2019",
month = "7",
doi = "10.1016/j.envres.2019.03.068",
language = "English",
volume = "174",
pages = "24--34",
journal = "Environmental Research",
issn = "0013-9351",
publisher = "Academic Press Inc.",

}

Comparison of model estimates from an intra-city land use regression model with a national satellite-LUR and a regional Bayesian Maximum Entropy model, in estimating NO 2 for a birth cohort in Sydney, Australia. / Cowie, Christine T.; Garden, Frances; Jegasothy, Edward; Knibbs, Luke D.; Hanigan, Ivan; Morley, David; Hansell, Anna; Hoek, Gerard; Marks, Guy B.

In: Environmental Research, Vol. 174, 07.2019, p. 24-34.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Comparison of model estimates from an intra-city land use regression model with a national satellite-LUR and a regional Bayesian Maximum Entropy model, in estimating NO 2 for a birth cohort in Sydney, Australia

AU - Cowie, Christine T.

AU - Garden, Frances

AU - Jegasothy, Edward

AU - Knibbs, Luke D.

AU - Hanigan, Ivan

AU - Morley, David

AU - Hansell, Anna

AU - Hoek, Gerard

AU - Marks, Guy B.

PY - 2019/7

Y1 - 2019/7

N2 - Background: Methods for estimating air pollutant exposures for epidemiological studies are becoming more complex in an effort to minimise exposure error and its associated bias. While land use regression (LUR) modelling is now an established method, there has been little comparison between LUR and other recent, more complex estimation methods. Our aim was to develop a LUR model to estimate intra-city exposures to nitrogen dioxide (NO 2 ) for a Sydney cohort, and to compare those with estimates from a national satellite-based LUR model (Sat-LUR) and a regional Bayesian Maximum Entropy (BME) model. Methods: Satellite-based LUR and BME estimates were obtained using existing models. We used methods consistent with the European Study of Cohorts for Air Pollution Effects (ESCAPE) methodology to develop LUR models for NO 2 and NOx. We deployed 46 Ogawa passive samplers across western Sydney during 2013/2014 and acquired data on land use, population density, and traffic volumes for the study area. Annual average NO 2 concentrations for 2013 were estimated for 947 addresses in the study area using the three models: standard LUR, Sat-LUR and a BME model. Agreement between the estimates from the three models was assessed using interclass correlation coefficient (ICC), Bland-Altman methods and correlation analysis (CC). Results: The NO 2 LUR model predicted 84% of spatial variability in annual mean NO 2 (RMSE: 1.2 ppb; cross-validated R 2 : 0.82) with predictors of major roads, population and dwelling density, heavy traffic and commercial land use. A separate model was developed that captured 92% of variability in NOx (RMSE 2.3 ppb; cross-validated R 2 : 0.90). The annual average NO 2 concentrations were 7.31 ppb (SD: 1.91), 7.01 ppb (SD: 1.92) and 7.90 ppb (SD: 1.85), for the LUR, Sat-LUR and BME models respectively. Comparing the standard LUR with Sat-LUR NO 2 cohort estimates, the mean estimates from the LUR were 4% higher than the Sat-LUR estimates, and the ICC was 0.73. The Pearson's correlation coefficients (CC) for the LUR vs Sat-LUR values were r = 0.73 (log-transformed data) and r = 0.69 (untransformed data). Comparison of the NO 2 cohort estimates from the LUR model with the BME blended model indicated that the LUR mean estimates were 8% lower than the BME estimates. The ICC for the LUR vs BME estimates was 0.73. The CC for the logged LUR vs BME estimates was r = 0.73 and for the unlogged estimates was r = 0.69. Conclusions: Our LUR models explained a high degree of spatial variability in annual mean NO 2 and NOx in western Sydney. The results indicate very good agreement between the intra-city LUR, national-scale sat-LUR, and regional BME models for estimating NO 2 for a cohort of children residing in Sydney, despite the different data inputs and differences in spatial scales of the models, providing confidence in their use in epidemiological studies.

AB - Background: Methods for estimating air pollutant exposures for epidemiological studies are becoming more complex in an effort to minimise exposure error and its associated bias. While land use regression (LUR) modelling is now an established method, there has been little comparison between LUR and other recent, more complex estimation methods. Our aim was to develop a LUR model to estimate intra-city exposures to nitrogen dioxide (NO 2 ) for a Sydney cohort, and to compare those with estimates from a national satellite-based LUR model (Sat-LUR) and a regional Bayesian Maximum Entropy (BME) model. Methods: Satellite-based LUR and BME estimates were obtained using existing models. We used methods consistent with the European Study of Cohorts for Air Pollution Effects (ESCAPE) methodology to develop LUR models for NO 2 and NOx. We deployed 46 Ogawa passive samplers across western Sydney during 2013/2014 and acquired data on land use, population density, and traffic volumes for the study area. Annual average NO 2 concentrations for 2013 were estimated for 947 addresses in the study area using the three models: standard LUR, Sat-LUR and a BME model. Agreement between the estimates from the three models was assessed using interclass correlation coefficient (ICC), Bland-Altman methods and correlation analysis (CC). Results: The NO 2 LUR model predicted 84% of spatial variability in annual mean NO 2 (RMSE: 1.2 ppb; cross-validated R 2 : 0.82) with predictors of major roads, population and dwelling density, heavy traffic and commercial land use. A separate model was developed that captured 92% of variability in NOx (RMSE 2.3 ppb; cross-validated R 2 : 0.90). The annual average NO 2 concentrations were 7.31 ppb (SD: 1.91), 7.01 ppb (SD: 1.92) and 7.90 ppb (SD: 1.85), for the LUR, Sat-LUR and BME models respectively. Comparing the standard LUR with Sat-LUR NO 2 cohort estimates, the mean estimates from the LUR were 4% higher than the Sat-LUR estimates, and the ICC was 0.73. The Pearson's correlation coefficients (CC) for the LUR vs Sat-LUR values were r = 0.73 (log-transformed data) and r = 0.69 (untransformed data). Comparison of the NO 2 cohort estimates from the LUR model with the BME blended model indicated that the LUR mean estimates were 8% lower than the BME estimates. The ICC for the LUR vs BME estimates was 0.73. The CC for the logged LUR vs BME estimates was r = 0.73 and for the unlogged estimates was r = 0.69. Conclusions: Our LUR models explained a high degree of spatial variability in annual mean NO 2 and NOx in western Sydney. The results indicate very good agreement between the intra-city LUR, national-scale sat-LUR, and regional BME models for estimating NO 2 for a cohort of children residing in Sydney, despite the different data inputs and differences in spatial scales of the models, providing confidence in their use in epidemiological studies.

KW - Air pollution

KW - Bayesian

KW - Comparison

KW - Exposure assessment

KW - Land use regression

KW - Satellite

UR - http://www.scopus.com/inward/record.url?scp=85064567291&partnerID=8YFLogxK

UR - http://www.mendeley.com/research/comparison-model-estimates-intracity-land-regression-model-national-satellitelur-regional-bayesian-m

U2 - 10.1016/j.envres.2019.03.068

DO - 10.1016/j.envres.2019.03.068

M3 - Article

VL - 174

SP - 24

EP - 34

JO - Environmental Research

T2 - Environmental Research

JF - Environmental Research

SN - 0013-9351

ER -