A model for external drift kriging with uncertain covariates applied to air quality measurements and dispersion model output

Kassteele, Jan van de; Stein, Alfred

doi:10.1002/env.771

Cited by 23 publications

(13 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Empirical models may give accurate results, but are restricted to the conditions under which they are developed (Manders, Schaap and Hoogerbrugge, 2009). Assimilating monitoring station observations and CTM output, with appropriate bias adjustments, has been shown to provide improvements over using either data source alone (van de Kassteele and Stein, 2006;Denby et al, 2008;Candiani et al, 2013;Hamm et al, 2015). In such settings, the CTM output enters as a model covariate and the measured station observations are the re- sponse.…”

mentioning

confidence: 99%

Nonseparable dynamic nearest neighbor Gaussian process models for large spatio-temporal data with an application to particulate matter analysis

et al. 2016

View full text Add to dashboard Cite

Particulate matter (PM) is a class of malicious environmental pollutants known to be detrimental to human health. Regulatory efforts aimed at curbing PM levels in different countries often require high resolution space–time maps that can identify red-flag regions exceeding statutory concentration limits. Continuous spatio-temporal Gaussian Process (GP) models can deliver maps depicting predicted PM levels and quantify predictive uncertainty. However, GP-based approaches are usually thwarted by computational challenges posed by large datasets. We construct a novel class of scalable Dynamic Nearest Neighbor Gaussian Process (DNNGP) models that can provide a sparse approximation to any spatio-temporal GP (e.g., with nonseparable covariance structures). The DNNGP we develop here can be used as a sparsity-inducing prior for spatio-temporal random effects in any Bayesian hierarchical model to deliver full posterior inference. Storage and memory requirements for a DNNGP model are linear in the size of the dataset, thereby delivering massive scalability without sacrificing inferential richness. Extensive numerical studies reveal that the DNNGP provides substantially superior approximations to the underlying process than low-rank approximations. Finally, we use the DNNGP to analyze a massive air quality dataset to substantially improve predictions of PM levels across Europe in conjunction with the LOTOS-EUROS chemistry transport models (CTMs).

show abstract

mentioning

confidence: 99%

Nonseparable dynamic nearest neighbor Gaussian process models for large spatio-temporal data with an application to particulate matter analysis

et al. 2016

View full text Add to dashboard Cite

show abstract

“…Measurements made with instruments at monitoring stations are considered authoritative; however, these observations are often too sparse to deliver regional maps at sufficient resolution to assess progress with mitigation strategies and for monitoring compliance. One solution is to couple spatially sparse monitoring station observations with spatially complete chemistry transport model (CTM) output, (see, e.g., van de Kassteele and Stein 2006;Denby, Schaap, Segers, Builtjes, and Horalek 2008;Candiani, Carnevale, Finzi, Pisoni, and Volta 2013). In such settings, monitoring station observations serve as a regression model outcome with CTM output set as a predictor.…”

Section: Analysis Of Air Pollution Datamentioning

confidence: 99%

Bayesian spatially varying coefficient models in the spBayes R package

Finley

Banerjee

2020

Environmental Modelling & Software

View full text Add to dashboard Cite

This paper describes and illustrates new functionality for fitting spatially varying coefficients models in the spBayes (version 0.4-2) R package. The new spSVC function uses a computationally efficient Markov chain Monte Carlo algorithm and extends current sp-Bayes functions, that fit only space-varying intercept regression models, to fit independent or multivariate Gaussian process random effects for any set of columns in the regression design matrix. Newly added OpenMP parallelization options for spSVC are discussed and illustrated, as well as helper functions for joint and point-wise prediction and model fit diagnostics. The utility of the proposed models is illustrated using a PM 10 analysis over central Europe.

show abstract

“…There are few and recent applications of geostatistical methods on transportation data. It can be observed that most available papers refer to traffic engineering studies (Ciuffo et al 2011;Mazzella et al 2011;Zou et al 2012), vehicle emission gases (Pearce et al 2009;Kassteele and Velders 2006;Kassteele and Stein 2006), and traffic injuries (Gundogdu 2014;Manepalli and Bham 2011;Molla et al 2014). Although in the 1990s, there were some studies in the literature on travel demand forecasting with applications of estimation techniques related to kriging (Ickstadt et al 1998), this approach is still recent and has not yet been sufficiently investigated in the line of research of transportation planning (Yoon et al 2014;Chen et al 2015).…”

Section: Introductionmentioning

confidence: 99%

A Conjoint Approach of Spatial Statistics and a Traditional Method for Travel Mode Choice Issues

Lindner

Pitombo

2017

J geovis spat anal

View full text Add to dashboard Cite

Conventional analysis of transportation demand is usually carried out using socioeconomic, travel, and land use attributes. Despite the effectiveness on travel demand forecasting, it is important to recognize that alternative approaches have been developed in recent years. Traditional methods, besides considering different explanatory variables, are appropriate to make estimates exclusively on previously surveyed households. On the other hand, recent studies have addressed spatial statistical concerns in the field of travel demand forecasting. The aim of this paper is to spatially estimate motorized travel mode choice probabilities in a continuous map using an Origin-Destination Survey database, conducted in the São Paulo Metropolitan Area in Brazil in 2007. Values were estimated in both sampled and non-sampled coordinates. This paper proposes a conjoint approach that combines the traditional procedure of travel demand forecasting (multiple logistic regression) with a spatial statistical method (ordinary kriging). A comparison is made with the one-step spatial method-indicator kriging (IK). Conjoint studies of spatial statistics and traditional methods are thriving in transportation analysis, giving rise to a travel mode choice surface in a confirmatory way. It is concluded that the proposed method can be used for future predictions of travel mode choices, unlike IK.

show abstract

A model for external drift kriging with uncertain covariates applied to air quality measurements and dispersion model output

Cited by 23 publications

References 12 publications

Nonseparable dynamic nearest neighbor Gaussian process models for large spatio-temporal data with an application to particulate matter analysis

Nonseparable dynamic nearest neighbor Gaussian process models for large spatio-temporal data with an application to particulate matter analysis

Bayesian spatially varying coefficient models in the spBayes R package

A Conjoint Approach of Spatial Statistics and a Traditional Method for Travel Mode Choice Issues

Contact Info

Product

Resources

About