Advanced

Reduced-rank spatio-temporal modeling of air pollution concentrations in the Multi-Ethnic Study of Atherosclerosis and Air Pollution

Olives, Casey; Sheppard, Lianne; Lindström, Johan LU ; Sampson, Paul D; Kaufman, Joel D and Szpiro, Adam A (2014) In Annals of Applied Statistics 8(4). p.2509-2537
Abstract
There is growing evidence in the epidemiologic literature of the relationship between air pollution and adverse health outcomes. Prediction of individual air pollution exposure in the Environmental Protection Agency (EPA) funded Multi-Ethnic Study of Atheroscelerosis and Air Pollution (MESA Air) study relies on a exible spatio-temporal prediction model that integrates land-use regression with kriging to account for spatial dependence in pollutant concentrations. Temporal variability is captured using temporal trends estimated via modified singular value decomposition and temporally varying spatial residuals. This model utilizes monitoring data from existing regulatory networks and supplementary MESA Air monitoring data to predict... (More)
There is growing evidence in the epidemiologic literature of the relationship between air pollution and adverse health outcomes. Prediction of individual air pollution exposure in the Environmental Protection Agency (EPA) funded Multi-Ethnic Study of Atheroscelerosis and Air Pollution (MESA Air) study relies on a exible spatio-temporal prediction model that integrates land-use regression with kriging to account for spatial dependence in pollutant concentrations. Temporal variability is captured using temporal trends estimated via modified singular value decomposition and temporally varying spatial residuals. This model utilizes monitoring data from existing regulatory networks and supplementary MESA Air monitoring data to predict concentrations for individual cohort members.



In general, spatio-temporal models are limited in their efficacy for large datasets due to computational intractability. We develop reduced-rank versions of the MESA Air spatio-temporal model. To do so, we apply low-rank kriging to account for spatial variation in the mean process and discuss the limitations of this approach. As an alternative, we represent spatial variation using thin plate regression splines. We compare the performance of the outlined models using EPA and MESA Air monitoring data for predicting concentrations of

oxides of nitrogen (NOx) - a pollutant of primary interest in MESA Air - in the Los Angeles metropolitan area via cross-validated R2.



Our findings suggest that use of reduced-rank models can improve computational eciency in certain cases. Low-rank kriging and thin plate regression splines were competitive across the formulations considered, although TPRS appeared to be more robust in some settings. (Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Contribution to journal
publication status
published
subject
in
Annals of Applied Statistics
volume
8
issue
4
pages
2509 - 2537
publisher
Institute of Mathematical Statistics
external identifiers
  • wos:000347530200028
  • scopus:84919461179
ISSN
1932-6157
DOI
10.1214/14-AOAS786
project
BECC
language
English
LU publication?
yes
id
ff860524-a54f-4fea-9c50-a64f7e820d7a (old id 4730261)
alternative location
http://www.e-publications.org/ims/submission/AOAS/user/submissionFile/17155?confirm=66f14426
date added to LUP
2014-10-27 13:13:43
date last changed
2017-01-01 04:16:58
@article{ff860524-a54f-4fea-9c50-a64f7e820d7a,
  abstract     = {There is growing evidence in the epidemiologic literature of the relationship between air pollution and adverse health outcomes. Prediction of individual air pollution exposure in the Environmental Protection Agency (EPA) funded Multi-Ethnic Study of Atheroscelerosis and Air Pollution (MESA Air) study relies on a exible spatio-temporal prediction model that integrates land-use regression with kriging to account for spatial dependence in pollutant concentrations. Temporal variability is captured using temporal trends estimated via modified singular value decomposition and temporally varying spatial residuals. This model utilizes monitoring data from existing regulatory networks and supplementary MESA Air monitoring data to predict concentrations for individual cohort members.<br/><br>
<br/><br>
In general, spatio-temporal models are limited in their efficacy for large datasets due to computational intractability. We develop reduced-rank versions of the MESA Air spatio-temporal model. To do so, we apply low-rank kriging to account for spatial variation in the mean process and discuss the limitations of this approach. As an alternative, we represent spatial variation using thin plate regression splines. We compare the performance of the outlined models using EPA and MESA Air monitoring data for predicting concentrations of<br/><br>
oxides of nitrogen (NOx) - a pollutant of primary interest in MESA Air - in the Los Angeles metropolitan area via cross-validated R2.<br/><br>
<br/><br>
Our findings suggest that use of reduced-rank models can improve computational eciency in certain cases. Low-rank kriging and thin plate regression splines were competitive across the formulations considered, although TPRS appeared to be more robust in some settings.},
  author       = {Olives, Casey and Sheppard, Lianne and Lindström, Johan and Sampson, Paul D and Kaufman, Joel D and Szpiro, Adam A},
  issn         = {1932-6157},
  language     = {eng},
  number       = {4},
  pages        = {2509--2537},
  publisher    = {Institute of Mathematical Statistics},
  series       = {Annals of Applied Statistics},
  title        = {Reduced-rank spatio-temporal modeling of air pollution concentrations in the Multi-Ethnic Study of Atherosclerosis and Air Pollution},
  url          = {http://dx.doi.org/10.1214/14-AOAS786},
  volume       = {8},
  year         = {2014},
}