Design of an optimum air monitoring network for exposure assessments

Nonlinear programming techniques are frequently used to design optimum monitoring networks. These mathematically rigorous techniques are difficult to implement or cumbersome when considering other design criteria. This paper presents a more pragmatic approach to the design of an optimal monitoring network to estimate human exposure to hazardous air pollutants. In this approach, an air quality simulation model is used to produce representative air quality patterns, which are then combined with population patterns to obtain typical exposure patterns. These combined patterns are used to determine ‘figures of merit’ for each potential monitoring site, which are used to identify and rank the most favorable sites. The spatial covariance structure of the air quality patterns is used to draw a ‘sphere of influence’ around each site to identify and eliminate redundant monitoring sites. This procedure determines the minimum number of sites required to achieve the desired spatial coverage. This methodology was used to design an optimal ambient air monitoring network for assessing population exposure to hazardous pollutants in the southeastern Ohio River valley.     

Spatial Modeling for Air Pollution Monitoring Network Design: Example of Residential Woodsmoke

Abstract

The purpose of this paper is to demonstrate how to develop an air pollution monitoring network to characterize small-area spatial contrasts in ambient air pollution concentrations. Using residential woodburning emissions as our case study, this paper reports on the first three stages of a four-stage protocol to measure, estimate, and validate ambient residential woodsmoke emissions in Vancouver, British Columbia. The first step is to develop an initial winter nighttime woodsmoke emissions surface using inverse-distance weighting of emissions information from consumer woodburning surveys and property assessment data. Second, fireplace density and a compound topo-graphic index based on hydrological flow regimes are used to enhance the emissions surface. Third, the spatial variation of the surface is used in a location-allocation algorithm to design a network of samplers for the woodsmoke tracer compound levoglucosan and fine particulate matter. Measurements at these network sites are then used in the fourth stage of the protocol (not presented here): a mobile sampling campaign aimed at developing a high-resolution surface of woodsmoke concentrations for exposure assignment in health effects studies. Overall the results show that relatively simple data inputs and spatial analysis can be effective in capturing the spatial variability of ambient air pollution emissions and concentrations.     

Sampled Monte Carlo uncertainty analysis for photochemical grid models

This study reports on the development and testing of a method of quantifying the uncertainties in concentration predictions by a complex photochemical grid model (PGM), using a modification of the basic Monte Carlo method (MCM). The computationally intensive aspects of applying a full MCM to hundreds of PGM inputs and model parameters is replaced by a highly restricted sampling approach that exploits the spatial persistence found in predicted concentration fields. The sampling approach to the MCM is being explored as an efficient approach to assess the uncertainty in the differences in predicted maximum ozone concentration between base case and control scenarios. The MCM is applied to several dozen surface cells, with the goal of sampling the spatial pattern of uncertainty in the PGM-predicted differences in surface ozone concentration fields between a pair of base and control scenarios. The uncertainty in model inputs and parameters is simulated using several types of stochastic models. These stochastic models are driven using Latin hypercube sampling (LHS) to generate a non-redundant ensemble of alternative model inputs. Preliminary testing of the sampled MCM approach was conducted using the UAM-IV PGM on the New York ozone attainment modeling domain for the 6–8 July 1988 ozone episode. One hundred alternative concentration estimates were generated for a base scenario and for control scenarios representing 50%, 10% and 5% reduction of NO_x emissions. The upper and lower bounds of the concentration difference ensemble that define a 95% confidence range were spatially interpolated from 27 monitoring sites to the full (surface) modeling domain, using the field of zero uncertainty (ZU) concentration differences. For the 50% NO_x control scenario, predicted increases in peak ozone concentration smaller than 20 ppb were generally not significant from zero. By contrast, predicted decreases in peak ozone greater than 10 ppb were usually significant. For a control scenario with a small 5% NO_x reduction, predicted concentration differences and confidence intervals were much smaller, but predicted changes in peak ozone were significant at a number of sample cells.     

Spatial modeling for air pollution monitoring network design: example of residential woodsmoke

The purpose of this paper is to demonstrate how to develop an air pollution monitoring network to characterize small-area spatial contrasts in ambient air pollution concentrations. Using residential woodburning emissions as our case study, this paper reports on the first three stages of a four-stage protocol to measure, estimate, and validate ambient residential woodsmoke emissions in Vancouver, British Columbia. The first step is to develop an initial winter nighttime woodsmoke emissions surface using inverse-distance weighting of emissions information from consumer woodburning surveys and property assessment data. Second, fireplace density and a compound topographic index based on hydrological flow regimes are used to enhance the emissions surface. Third, the spatial variation of the surface is used in a location-allocation algorithm to design a network of samplers for the woodsmoke tracer compound levoglucosan and fine particulate matter. Measurements at these network sites are then used in the fourth stage of the protocol (not presented here): a mobile sampling campaign aimed at developing a high-resolution surface of woodsmoke concentrations for exposure assignment in health effects studies. Overall the results show that relatively simple data inputs and spatial analysis can be effective in capturing the spatial variability of ambient air pollution emissions and concentrations.     

Characterization of Spatially Homogeneous Regions Based on Temporal Patterns of Fine Particulate Matter in the Continental United States

Abstract

Statistical analyses of time-series or spatial data have been widely used to investigate the behavior of ambient air pollutants. Because air pollution data are generally collected in a wide area of interest over a relatively long period, such analyses should take into account both spatial and temporal characteristics. The objective of this study is 2-fold: (1) to identify an efficient way to characterize the spatial variations of fine particulate matter (PM_2.5) concentrations based solely upon their temporal patterns, and (2) to analyze the temporal and seasonal patterns of PM_2.5 concentrations in spatially homogenous regions. This study used 24-hr average PM_2.5 concentrations measured every third day during a period between 2001 and 2005 at 522 monitoring sites in the continental United States. A k-means clustering algorithm using the correlation distance was used to investigate the similarity in patterns between temporal profiles observed at the monitoring sites. A k-means clustering analysis produced six clusters of sites with distinct temporal patterns that were able to identify and characterize spatially homogeneous regions of the United States. The study also presents a rotated principal component analysis (RPCA) that has been used for characterizing spatial patterns of air pollution and discusses the difference between the clustering algorithm and RPCA.     

A holistic approach for optimal design of air quality monitoring network expansion in an urban area

This paper presents an objective methodology for determining the optimum number of ambient air quality stations in a monitoring network. The methodology integrates the multiple-criteria method with the spatial correlation technique. The pollutant concentration and population exposure data are used in this methodology in different ways. In the first stage, the Fuzzy Analytic Hierarchy Process (FAHP) with triangular fuzzy numbers (TFNs) is used to identify the most desirable monitoring locations. The network configuration is then determined on the basis of the concept of sphere of influences (SOIs). The SOIs are dictated by a predetermined cutoff value (r_c) in the spatial correlation coefficients (r) between the pollutant concentrations at the monitoring stations identified from first step and the corresponding concentrations at neighboring locations in the region. Finally, the optimal station locations are ranked by using combined utility scores gained from the first and second steps. The expansion of air quality monitoring network of Riyadh city in Saudi Arabia is used as a case study to demonstrate the proposed methodology.     

Monitoring of heavy metal concentrations in home outdoor air using moss bags

One monitoring station is insufficient to characterize the high spatial variation of traffic-related heavy metals within cities. We tested moss bags (Hylocomium splendens), deployed in a dense network, for the monitoring of metals in outdoor air and characterized metals’ long-term spatial distribution and its determinants in Girona, Spain. Mosses were exposed outside 23 homes for two months; NO₂ was monitored for comparison. Metals were not highly correlated with NO₂ and showed higher spatial variation than NO₂. Regression models explained 61-85% of Cu, Cr, Mo, Pb, Sb, Sn, and Zn and 72% of NO₂ variability. Metals were strongly associated with the number of bus lines in the nearest street. Heavy metals are an alternative traffic-marker to NO₂ given their toxicological relevance, stronger association with local traffic and higher spatial variability. Monitoring heavy metals with mosses is appealing, particularly for long-term exposure assessment, as mosses can remain on site many months without maintenance.     

Comparison of sulfate and nitrate particle mass concentrations measured by IMPROVE and the CDN

The 1977 and 1990 Amendments to the Clean Air Act call for visibility and atmospheric deposition monitoring throughout the United States. We compare sulfate and nitrate particle mass concentrations measured by two regional air quality networks, the Interagency Monitoring of PROtected Visual Environments (IMPROVE) network and the Clean Air Status and Trends Network (CASTNet), or CASTNet Deposition Network (CDN). The intent of this comparison is to quantify bias that may be introduced from differences in the respective network's sampling protocols. A number of sampling protocol differences exist between the two networks that may lead to sampling bias, particularly for particle NO₃⁻. Observed differences between particle SO₄²⁻ mass concentrations reported by the two monitoring networks are generally small, yet statistically significant at many comparison sites. Differences between particle NO₃⁻ mass concentrations are substantial, statistically significant at nearly all comparison sites, and the bias magnitude varies by geographic region. Differences in particle NO₃⁻, based on data from monitoring sites selected for this comparison, are 40% in the west, 56% in the interior desert/mountain region, and −9% in the east, expressed as the IMPROVE mean subtracted from the CDN mean, as a percent of the IMPROVE mean. Comparisons are made using data from 23 locations where monitoring sites from IMPROVE and CDN are within approximately 50 km.     

Exposure assessment of air pollutants: a review on spatial heterogeneity and indoor/outdoor/personal exposure to suspended particulate matter,nitrogen dioxide and ozone

This review describes databases of small-scale spatial variations and indoor, outdoor and personal measurements of air pollutants with the main focus on suspended particulate matter, and to a lesser extent, nitrogen dioxide and photochemical pollutants. The basic definitions and concepts of an exposure measurement are introduced as well as some study design considerations and implications of imprecise exposure measurements. Suspended particulate matter is complex with respect to particle size distributions, the chemical composition and its sources. With respect to small-scale spatial variations in urban areas, largest variations occur in the ultrafine (<0.1 μm) and the coarse mode (PM_10–2.5, resuspended dust). Secondary aerosols which contribute to the accumulation mode (0.1–2 μm) show quite homogenous spatial distribution. In general, small-scale spatial variations of PM_2.5 were described to be smaller than the spatial variations of PM₁₀. Recent studies in outdoor air show that ultrafine particle number counts have large spatial variations and that they are not well correlated to mass data. Sources of indoor particles are from outdoors and some specific indoor sources such as smoking and cooking for fine particles or moving of people (resuspension of dust) for coarse particles. The relationships between indoor, outdoor and personal levels are complex. The finer the particle size, the better becomes the correlation between indoor, outdoor and personal levels. Furthermore, correlations between these parameters are better in longitudinal analyses than in cross-sectional analyses. For NO₂ and O₃, the air chemistry is important. Both have considerable small-scale spatial variations within urban areas. In the absence of indoor sources such as gas appliances, NO₂ indoor/outdoor relationships are strong. For ozone, indoor levels are quite small. The study hypothesis largely determines the choice of a specific concept in exposure assessment, i.e. whether personal sampling is needed or if ambient monitoring is sufficient. Careful evaluation of the validity and improvements in precision of an exposure measure reduce error in the measurements and bias in the exposure–effect relationship.     

Spatial variation of airborne pollutant concentrations in Brisbane,Australia and its potential impact on population exposure assessment

Spread of air pollution sources and non-uniform mixing conditions in urban or regional air sheds often result in spatial variation of pollutant concentrations over different parts of the air sheds. A comprehensive understanding of this variation of concentrations is imperative for informed planning, monitoring and assessment in a range of critical areas including assessment of monitoring network efficiency or assessment of population exposure variation as a function of the location in the city. The aims of this work were to study the citywide variability of pollutants as measured by “urban background” type monitoring stations and to interpret the results in relation to the applicability of the data to population exposure assessments and the network efficiency. A comparison between ambient concentrations of NO_x, ozone and PM₁₀ was made for three stations in the Brisbane air shed network. The best correlated between the three stations were ozone concentrations followed by NO_x concentration, with the worst correlations observed for PM₁₀. With a few exceptions correlations of all pollutants between the stations were statistically significant. Marginally better were the correlations for the lower concentrations of pollutants that represent urban background, over the correlations for higher concentrations, representing peak values. Implications of these findings on application of the monitoring data to air-quality management, as well as the need for further investigations has been discussed.     

Assessing the spatial distribution of nitrogen dioxide in London,Ontario

Land use regression (LUR) models have been widely used to characterize the spatial distribution of urban air pollution and estimate exposure in epidemiologic studies. However, spatial patterns of air pollution vary greatly between cities due to local source type and distribution. London, Ontario, Canada, is a medium-sized city with relatively few and isolated industrial point sources, which allowed the study to focus on the contribution of different transportation sectors to urban air pollution. This study used LUR models to estimate the spatial distribution of nitrogen dioxide (NO₂) and to identify local sources influencing NO₂ concentrations in London, ON. Passive air sampling was conducted at 50 locations throughout London over a 2-week period in May–June 2010. NO₂ concentrations at the monitored locations ranged from 2.8 to 8.9 ppb, with a median of 5.2 ppb. Industrial land use, dwelling density, distance to highway, traffic density, and length of railways were significant predictors of NO₂ concentrations in the final LUR model, which explained 78% of NO₂ variability in London. Traffic and dwelling density explained most of the variation in NO₂ concentrations, which is consistent with LUR models developed in other Canadian cities. We also observed the importance of local characteristics. Specifically, 17% of the variation was explained by distance to highways, which included the impacts of heavily traveled corridors transecting the southern periphery of the city. Two large railway yards and railway lines throughout central areas of the city explained 9% of NO₂ variability. These results confirm the importance of traditional LUR variables and highlight the importance of including a broader array of local sources in LUR modeling. Finally, future analyses will use the model developed in this study to investigate the association between ambient air pollution and cardiovascular disease outcomes, including plaque burden, cholesterol, and hypertension.

Implications: Monitoring and modeling of NO₂ throughout the city of London represents an important step toward assessing air pollution health effects in a mid-sized Canadian city. The study supports the introduction of railways to LUR modeling of NO₂. Railways explained approximately 9% of the variability in ambient NO₂ concentrations in London, which suggests that local sources captured by land-use indicators may contribute to the efficacy of LUR models. These findings provide insights relevant to other medium and smaller sized cities with similar land use and transportation infrastructure. Furthermore, London is a central hub for medical research and treatment in southwestern Ontario, with facilities such as the Robarts Research Institute, London Regional Cancer Program (LRCP), and Stroke Prevention & Atherosclerosis Research Centre (SPARC). The models developed in this study will provide estimates of exposure for future analyses examining air pollution health effects in this data-rich population.     

A land use regression model for predicting ambient volatile organic compound concentrations in Toronto,Canada

More than 25 studies have employed land use regression (LUR) models to estimate nitrogen oxides and to a lesser extent particulate matter indicators, but these methods have been less commonly applied to ambient concentrations of volatile organic compounds (VOCs). Some VOCs have high plausibility as sources of health effects and others are specific indicators of motor vehicle exhaust. We used LUR models to estimate spatial variability of VOCs in Toronto, Canada. Benzene, n-hexane and total hydrocarbons (THC) were measured from July 25 to August 9, 2006 at 50 locations using the TraceAir organic vapor monitors. Nitrogen dioxide (NO₂) was also sampled to assess its spatial pattern agreement with VOC exposures. Buffers for land use, population density, traffic density, physical geography, and remote sensing measures of greenness and surface brightness were also tested. The remote sensing measures have the highest correlations with VOCs and NO₂ levels (i.e., explains >36% of the variance). Our regression models explain 66–68% of the variance in the spatial distribution of VOCs, compared to 81% for the NO₂ model. The ranks of agreement between various VOCs range from 48 to 63% and increases substantially – up to 75% – for the top and bottom quartile groups. Agreements between NO₂ and VOCs are much smaller with an average rank of 36%. Future epidemiologic studies may therefore benefit from using VOCs as potential toxic agents for traffic-related pollutants.     

Disaggregation of nation-wide dynamic population exposure estimates in The Netherlands: Applications of activity-based transport models

Traditional exposure studies that link concentrations with population data do not always take into account the temporal and spatial variations in both concentrations and population density. In this paper we present an integrated model chain for the determination of nation-wide exposure estimates that incorporates temporally and spatially resolved information about people's location and activities (obtained from an activity-based transport model) and about ambient pollutant concentrations (obtained from a dispersion model). To the best of our knowledge, it is the first time that such an integrated exercise was successfully carried out in a fully operational modus for all models under consideration. The evaluation of population level exposure in The Netherlands to NO₂ at different time-periods, locations, for different subpopulations (gender, socio-economic status) and during different activities (residential, work, transport, shopping) is chosen as a case-study to point out the new features of this methodology. Results demonstrate that, by neglecting people's travel behaviour, total average exposure to NO₂ will be underestimated by 4% and hourly exposure results can be underestimated by more than 30%. A more detailed exposure analysis reveals the intra-day variations in exposure estimates and the presence of large exposure differences between different activities (traffic > work > shopping > home) and between subpopulations (men > women, low socio-economic class > high socio-economic class). This kind of exposure analysis, disaggregated by activities or by subpopulations, per time of day, provides useful insight and information for scientific and policy purposes. It demonstrates that policy measures, aimed at reducing the overall (average) exposure concentration of the population may impact in a different way depending on the time of day or the subgroup considered. From a scientific point of view, this new approach can be used to reduce exposure misclassification.     

Spatial characteristics of fine particulate matter: identifying representative monitoring locations in Seattle, Washington

This study investigates how PM2.5 varies spatially and how these spatial characteristics can be used to identify potential monitoring sites that are most representative of the overall ambient exposures to PM2.5 among susceptible populations in the Seattle, WA, area. Data collected at outdoor sites at the homes of participants of a large exposure assessment study were used in this study. Harvard impactors (HIs) were used at 40 outdoor sites throughout the Seattle metropolitan area. Up to six sites at a time were monitored for 10 consecutive 24-hr average periods. A fixed-effect analysis of variance (ANOVA) model that included date and location effects was used to analyze the spatial variability of outdoor PM2.5 concentrations. Both date and location effects were shown to be highly significant, explaining 92% of the variability in outdoor PM2.5 measurements. The day-to-day variability was 10 times higher than the spatial variability between sites. The site mean square was more than twice the error mean square, showing that differences between sites, while modest, are potentially an important contribution to measurement error. Variances of the model residuals and site effects were examined against spatial characteristics of the monitoring sites. The spatial characteristics included elevation, distance from arterials, and distance from major PM2.5 point sources. Results showed that the most representative PM2.5 sites were located at elevations of 80-120 m above sea level, and at distances of 100-300 m from the nearest arterial road. Location relative to industrial PM2.5 sources is not a significant predictor of residential outdoor PM2.5 measurements. Additionally, for sites to be representative of the average population exposures to PM2.5 among those highly susceptible to the health effects of PM2.5, areas of high elderly population density were considered. These representative spatial characteristics were used as multiple, overlapping criteria in a Geographic Information System (GIS) analysis to determine where the most representative sites are located.     

A prediction-based approach to modelling temporal and spatial variability of traffic-related air pollution in Montreal, Canada

Concentrations of traffic-related air pollution can be highly variable at the local scale and can have substantial seasonal variability. This study was designed to provide estimates of intra-urban concentrations of ambient nitrogen dioxide (NO₂) in Montreal, Canada, that would be used subsequently in health studies of chronic diseases and long-term exposures to traffic-related air pollution. We measured concentrations of NO₂ at 133 locations in Montreal with passive diffusion samplers in three seasons during 2005 and 2006. We then used land use regression, a proven statistical prediction method for describing spatial patterns of air pollution, to develop separate estimates of spatial variability across the city by regressing NO₂ against available land-use variables in each of these three periods. We also developed a “pooled” model across these sampling periods to provide an estimate of an annual average. Our modelling strategy was to develop a predictive model that maximized the model R². This strategy is different from other strategies whose goal is to identify causal relationships between predictors and concentrations of NO₂.Observed concentrations of NO₂ ranged from 2.6 ppb to 31.5 ppb, with mean values of 12.6 ppb in December 2005, 14.0 ppb in May 2006, and 8.9 ppb in August 2006. The greatest variability was observed during May. Concentrations of NO₂ were highest downtown and near major highways, and they were lowest in the western part of the city. Our pooled model explained approximately 80% of the variability in concentrations of NO₂. Although there were differences in concentrations of NO₂ between the three sampling periods, we found that the spatial variability did not vary significantly across the three sampling periods and that the pooled model was representative of mean annual spatial patterns.     

Co-occurrence patterns of gaseous air pollutant pairs at different minimum concentrations in the United States

The frequency of co-occurrences for SO₂NO₂, SO₂/O₃ and O₃/NO₂ at rural and remote monitoring sites in the United States was characterized for the months of May-September for the years 1978–1982. Minimum hourly concentrations of 0.03 and 0.05 ppm of each gas were used as the criteria for defining a ‘co-occurrence’. The objectives of this study were to:

• 1.(1) identify the types of co-occurrence patterns and their frequency;
• 2.(2) identify whether the frequency of hourly simultaneous co-occurrences increased substantially when the minimum concentration was lowered (e.g. from 0.05 to 0.03 ppm) for each pollutant; and
• 3.(3) determine whether the frequency of co-occurrences showed large year-to-year variation.

For all pollutant pairs and co-occurrence thresholds (i.e. 0.03 and 0.05 ppm), the frequency of daily and hourly co-occurrences was low for most sites. Year-to-year variability was found to be insignificant; most of the monitoring sites experienced co-occurrences of any type less than 12% of the 153 days. Based on our observations, researchers attempting to assess the potential effects of SO₂/NO₂, SO₂/O₃ and O₃/NO₂ in the United States should construct simulated exposure regimes so that

• 1.(1) hourly simultaneous and daily simultaneous-only co-occurrences are fairly rare and
• 2.(2) when co-occurrences are present, complex-sequential and sequential-only co-occurrence patterns predominate.

     

A Short-Term Diffusive Sampler for Nitrogen Dioxide Monitoring in Epidemiology

An automated timed exposure diffusive sampler (TEDS) for sampling nitrogen dioxide (NO₂) was developed for use in epidemiological studies. The TEDS sequentially exposes four passive sampling devices (PSD) by microprocessor controlled valves while a pump and air flow guide prevent sampler "starvation." Two TEDS units and two portable, real-time NO₂ monitors were tested for accuracy, precision, sensitivity, and linearity of response. The accuracy of the TEDS was within 10 percent of the calibrated NO₂ values, and precision was within 10 percent of the means of the measured values. The TEDS sensitivity was 20 to 30 ppb-hour for NO₂. Co-location of the TEDS with a chemiluminescent NOX monitor (EPA reference method) showed similar responses to ambient NO₂ (R2 = 0.9991). TEDS allows better time resolution than traditional diffusive samplers (i.e., Palmes tube) while sharing their ability to sample a variety of gases.     

Comparative evaluation of nitrogen oxides and ozone passive diffusion tubes for exposure studies

Passive diffusion tubes are recognised as a cost-effective sampling method for characterising the spatial variability, as well as the seasonal and annual trends, of NO₂ concentrations in urban areas. In addition, NO_X and O₃ passive diffusion tubes have been developed and deployed in urban and rural areas. Despite their many advantages (e.g. low operational and analysis cost, small size and no need for power supply), they have certain limitations mainly related to their accuracy and precision. In particular, the absorbent solution used, the length of the exposure period, the exact location and use of protective devices, and other environmental conditions (e.g. wind, ambient temperature and relative humidity) may have a significant impact on the performance of passive diffusion tubes. The aim of this study is to evaluate the performance of co-located NO₂, NO_X and O₃ diffusion tubes in an urban environment.A one-year passive sampling campaign was carried out in Birmingham (UK) for this purpose. NO₂, NO_X and O₃ diffusion tubes (including triplicate sets of each) were co-located at one urban background and two roadside permanent air quality monitoring stations equipped with standard gas analysers. In addition, meteorological data, such as wind speed and direction, ambient temperature and relative humidity, were obtained during the same period of time. A thorough QA/QC procedure, including storage and laboratory blanks was followed throughout the campaign. The analysis of results showed a very good agreement of NO₂ passive samplers with co-located chemiluminescence analysers, but substantial underestimations of total NO_X levels by the diffusion tubes. The O₃ diffusion sampler appeared to marginally overestimate the automatic UV analyser results, especially during warm weather periods.     

Identification of redundant air quality measurements through the use of principal component analysis

This study aims to show how principal component analysis (PCA) can be used to identify redundant measurements in air quality monitoring networks. The minimum number of air quality monitoring sites in Oporto Metropolitan Area (Oporto-MA) was evaluated using PCA and then compared to the one settled by the legislation. Nine sites, monitoring NO₂, O₃ and PM₁₀, were selected and the air pollutant concentrations were analysed from January 2003 to December 2005. PCA was applied to the data corresponding to the first two years that were divided into annual quarters to verify the persistence of the PCA results. The number of principal components (PCs) was selected by applying two criteria: Kaiser (PCs with eigenvalues greater than 1) and ODV₉₀ (PCs representing at least 90% of the original data variance). Each pollutant was analysed separately. The two criteria led to different results. Using Kaiser criterion for the eight analysed periods, two PCs were selected in: (i) five periods for O₃ and PM₁₀; and (ii) six periods for NO₂. These PCs had important contributions of the same groups of monitoring sites. The percentage of the original data variance contained in the selected PCs using this criterion was always below 90%. Thus, the results obtained using ODV₉₀ were considered with more confidence. Using this criterion, only five monitoring sites for NO₂, three for O₃ and seven for PM₁₀ were needed to characterize the region. The number of monitoring sites for NO₂ and O₃ was in agreement with what was established by the legislation. However, for PM₁₀, Oporto-MA needed two more monitoring sites. To validate PCA results, statistical models were determined to estimate air pollutant concentrations at removed monitoring sites using the concentrations measured at the remaining monitoring sites. These models were applied to a year's data. The good performance obtained by the models showed that the monitoring sites selected by the procedure presented in this study were enough to infer the air pollutant concentrations in the region defined by the initial monitoring sites. Additionally, the air pollutant analysers corresponding to the redundant measurements can be installed in non-monitored regions, allowing the enlargement of the air quality monitoring network.     

Cost-effective sampling network design for contaminant plume monitoring under general hydrogeological conditions

A new simulation-optimization methodology is developed for cost-effective sampling network design associated with long-term monitoring of large-scale contaminant plumes. The new methodology is similar in concept to the one presented by Reed et al. (Reed, P.M., Minsker, B.S., Valocchi, A.J., 2000a. Cost-effective long-term groundwater monitoring design using a genetic algorithm and global mass interpolation. Water Resour. Res. 36 (12), 3731-3741) in that an optimization model based on a genetic algorithm is coupled with a flow and transport simulator and a global mass estimator to search for optimal sampling strategies. However, this study introduces the first and second moments of a three-dimensional contaminant plume as new constraints in the optimization formulation, and demonstrates the proposed methodology through a real-world application. The new moment constraints significantly increase the accuracy of the plume interpolated from the sampled data relative to the plume simulated by the transport model. The plume interpolation approaches employed in this study are ordinary kriging (OK) and inverse distance weighting (IDW). The proposed methodology is applied to the monitoring of plume evolution during a pump-and-treat operation at a large field site. It is shown that potential cost savings up to 65.6% may be achieved without any significant loss of accuracy in mass and moment estimations. The IDW-based interpolation method is computationally more efficient than the OK-based method and results in more potential cost savings. However, the OK-based method leads to more accurate mass and moment estimations. A comparison of the sampling designs obtained with and without the moment constraints points to their importance in ensuring a robust long-term monitoring design that is both cost-effective and accurate in mass and moment estimations. Additional analysis demonstrates the sensitivity of the optimal sampling design to the various coefficients included in the objective function of the optimization model.