Search

In this study, PM2.5 concentrations together with the water-soluble ionic compounds and total elements were simultaneously measured at 16 sites in the city of Varese located in Northern Italy within a domain of 2 × 2 km2. The center point of this domain consisted of an existing urban air quality monitoring station. The representativeness of the monitoring station for PM2.5 mass and chemical composition was estimated using a methodology relying on statistical/geostatistical analyses. Source apportionment by means of the Chemical Mass Balance (CMB) receptor model was also performed to evaluate the spatial distribution of source contribution throughout the domain. Very high soluble fraction was found for Cd, Sb, K and V, indicating the anthropogenic origin of those elements. The geostatistical analysis/mapping showed that the monitoring station captured the spatial variation of PM2.5 and most of the anthropogenic originated elements, e.g., As, Cd and V, whereas it did not captured the spatial patterns of the ones originated from both natural and anthropogenic sources, e.g., Na, Ni, Pb, K, Zn, Fe, Cr, and Ti. The CMB source contribution estimations in the monitoring station were at least 25% different from many sites of the domain for PM2.5. The significant spatial variation in concentrations and source contribution estimates showed that the monitoring station could not be considered representative for the air quality monitoring studies with exposure assessment and source apportionment purposes in Varese.

Over the last few years there have been many activities to evaluate and use air quality measurements gathered by lower cost devices. This is especially intended to complement the coverage of official air quality measurement networks that deliver authoritative air quality data. Examples of such activities include the AirSensEUR project of the Joint Research Centre (JRC), luftdaten.info, or hackAIR. Combining the data from multiple sources remains a challenge for utilising the full potential of those developments.With this presentation we aim to introduce the development of an interoperable data platform that allows to integrate both authoritative as well as citizen science air quality measurements. Our presentation will cover especially the following aspects:Interoperability: For sharing the collected data and to avoid the creation of isolated data silos, it is important to use open interfaces and data encodings. In case of the AQSens project, this comprises the provision of INSPIRE-compliant Download Services based on the SensorThings API (STA) and Sensor Observation Service (SOS) standards of the Open Geospatial Consortium.Data analytics: Besides providing access to the raw data, different types of data analysis are necessary. On the one hand this comprises the validation of incoming citizen science data in conjunction with corresponding authoritative data sources. On the other hand, the aim is to provide a tool for further data analysis on top of the collected data. For this purpose we show, how the R programming language can be linked to the Sensor Web Server via a dedicated R package (sos4R).Data visualisation: Finally, for enabling the visual exploration of the collected data, a Web-based client application will be provided. This allows users to connect to the published air quality Data Download Services (in this case the OGC SensorThings API) and to request graph-based time series visualisations combining data from potentially different sources.In summary, our presentation will show how existing interoperability standards as well as Web technologies can be used for building a Cloud-ready data platform (i.e. relying on Docker) that enables the collection, management, analysis, and visualisation of both Citizen Science and authoritative air quality data.

The modified Target Diagram (MTD) was developed to evaluate the performance of low-cost sensors (LCS) for air quality monitoring in comparison with reference methods by reporting relative expanded uncertainty and its contributors. An MTD provides several pieces of information, including compliance with regulation, sources of error and how to diminish them, completeness and validity of LCS calibration etc. It allows the user to examine the effect of selecting different regression types and residual fitting on the LCS measurement uncertainty. The ordinary least squared regression with fitted residuals and dynamic between reference analyser uncertainty rather than constant ones yielded more realistic LCS measurement uncertainty compared to other options. The MTD is a fast visual tool to extract several pieces of information on evaluation of any candidate method against reference method.

For detection of benzene, a gas sensor system with metal oxide semiconductor (MOS) gas sensors using temperature-cycled operation (TCO) is presented. The system has been tested in two different laboratories at the concentration range from 0.5 up to 10 ppb. The system is equipped with three gas sensors and advanced temperature control and read-out electronics for the extraction of features from the TCO signals. A sensor model is used to describe the sensor response in dependence on the gas concentration. It is based on a linear differential surface reduction (DSR) at a low temperature phase, which is linked to an exponential growth of the sensor conductance. To compensate for cross interference to other gases, the DSR is measured at three different temperatures (200, 250, 300 ∘C) and the calculated features are put into a multilinear regression (partial least square regression – PLSR) for the quantification of benzene at both laboratories. In the tests with the first set-up, benzene was supplied in defined gas profiles in a continuous gas flow with variation of humidity and various interferents, e.g. toluene and carbon monoxide (CO). Depending on the gas background and interferents, the quantification accuracy is between ±0.2 and ±2 ppb. The second gas mixing system is based on a circulation of the carrier gas stream in a closed-loop control for the benzene concentration and other test gases based on continuously available reference measurements for benzene and other organic and inorganic compounds. In this system, a similar accuracy was achieved for low background contaminations and constant humidity; the benzene level could be quantified with an error of less than 0.5 ppb. The transfer of regression models for one laboratory to the other has been tested successfully.

EDXRF analysis is typically carried out under vacuum to reach the optimum performance, particularly for low Z elements. However, when analyzing PM10 loaded filters, the application of vacuum is responsible for the mass loss of volatile components preventing further analysis on the same sample. Alternatives include EDXRF analysis in helium (He) or air medium. The standardless EDXRF analysis using He and air analytical medium was evaluated against analysis under vacuum. The spectral peak intensities and quantification results of EDXRF using He and air medium were compared to the ones using vacuum. The intensities were found to be almost identical except for Cl, Mo and Ba. The quantification results were similar with a few exceptions. In addition to standardless analysis, the external calibration using standards analyzed by ICP-MS was also tested. The measurement uncertainty of standardless EDXRF analysis using He and air medium was estimated by a pre-established model equation and by a field study. The model equation was confirmed by the field study for many elements, except for Ca, Mg, Ni and Cr. The data quality objective (DQO) of the European Directive for Air Quality was met for Pb by the standardless EDXRF analysis using He and air medium, and for As by the external calibration. For Ni and Cd, none of the quantification methods could meet the DQO, most likely due to low loadings of those elements on the studied PM10 samples. It was demonstrated that analyzing air filters by the standardless EDXRF analysis using He and air medium is a valid alternative to vacuum.

In this work the performances of several field calibration methods for low-cost sensors, including linear/multi linear regression and supervised learning techniques, are compared. A cluster of either metal oxide or electrochemical sensors for nitrogen monoxide and carbon monoxide together with miniaturized infra-red carbon dioxide sensors was operated. Calibration was carried out during the two first weeks of evaluation against reference measurements. The accuracy of each regression method was evaluated on a five months field experiment at a semi-rural site using different indicators and techniques: orthogonal regression, target diagram, measurement uncertainty and drifts over time of sensor predictions. In addition to the analyses for ozone and nitrogen oxide already published in Part A [1], this work assessed if carbon monoxide sensors can reach the Data Quality Objective (DQOs) of 25% of uncertainty set in the European Air Quality Directive for indicative methods. As for ozone and nitrogen oxide, it was found for NO, CO and CO 2 that the best agreement between sensors and reference measurements was observed for supervised learning techniques compared to linear and multilinear regression.

Multi-element reference materials (ME-RM) were prepared with 28 elements on polytetrafluoroethylene (PTFE) filters utilizing a custom aerosol deposition system and certified multi-element solutions designed to mimic atmospheric particulate composition. ME-RM were then distributed to eight participating laboratories who provided measurement results via X-ray fluorescence (XRF), particle induced X-ray emission (PIXE), and inductively coupled plasma mass spectrometry (ICP-MS). Lab-consensus loadings computed following an international methodology agreed well with a priori ones for Na, Al, S, K, Ca, Ti, V, Cr, Mn, Fe, Co, Ni, As, Se, Rb, Sr, Cd, and Pb. The lab results were in good agreement with a priori reference loadings for S, K, Ca, V, Mn, Fe, Co, As, Se, Rb, Sr, Cd, and Pb. This work demonstrates the ability of the custom aerosol chamber to generate consistent reference filters that can be used for instrument calibration and quality control, interlaboratory comparison, and proficiency testing purposes, filling a much-needed gap hitherto lacking because of unavailability of suitable reference materials.

Eight institutes using 12 different instruments analyzed newly developed multi-element reference materials (RM) for atmospheric particulate matter (PM) measurements. These RM have the potential to fill a gap in the currently available quality assurance resources for element analysis of PM samples such as X-ray fluorescence and inductively-coupled plasma mass spectrometry. This study evaluates the performance of these new RM generated by the University of California, Davis. The methodological challenge was to determine the reference loadings on the RM. Gravimetry is the most robust method to determine the sample deposit mass but cannot be used for these RM because some solution components are volatile and result in unpredictable total mass loadings on the RM. Instead of using gravimetry, a single well-measured element, along with the assumption that the relative mass fractions in the solutions were maintained in the aerosol deposited on the filters, was used to determine the reference loadings on the RM. This assumption appears to be valid for most elements in the solutions; notable exceptions include volatile species such as chlorine and bromine. Results from the 12 different instruments in the inter-laboratory evaluation agreed very well with the reference loadings (adjusted R2 > 0.9 and slope between 0.7 and 1.3) for 17 of the 28 elements. In many cases, one or two instruments did not meet the performance criteria, which points to individual instrument calibration problems. For the 11 elements that did not perform as well, development work continues, and this intercomparison helped identify and fix a source of contamination in the system used to create the RM. Copyright © 2019 American Association for Aerosol Research

An inter-laboratory comparison study was conducted to measure elemental loadings on PM 10 samples, collected in Ispra, a regional background/rural site in Italy, using three different XRF (X-ray Fluorescence) methods, namely Epsilon 5 by linear calibration, Quant’X by the standardless analysis, and PIXE (Particle Induced X-ray Emission) with linear calibration. A subset of samples was also analyzed by ICP-MS (Inductively Coupled Plasma-Mass Spectrometry). Several metrics including method detection limits (MDLs), precision, bias from a NIST standard reference material (SRM 2783) quoted values, relative absolute difference, orthogonal regression and the ratio of the absolute difference between the methods to claimed uncertainty were used to compare the laboratories. The MDLs were found to be comparable for many elements. Precision estimates were less than 10% for the majority of the elements. Absolute biases from SRM 2783 remained less than 20% for the majority of certified elements. The regression results of PM 10 samples showed that the three XRF laboratories measured very similar mass loadings for S, K, Ti, Mn, Fe, Cu, Br, Sr and Pb with slopes within 20% of unity. The ICP-MS results confirmed the agreement and discrepancies between XRF laboratories for Al, K, Ca, Ti, V, Cu, Sr and Pb. The ICP-MS results are inconsistent with the XRF laboratories for Fe and Zn. The absolute differences between the XRF laboratories generally remained within their claimed uncertainties, showing a pattern generally consistent with the orthogonal regression results.

A new methodology to assess source apportionment model performance in intercomparison exercises, encompassing the preparation of real-world and synthetic datasets and the evaluation of the source apportionment results reported by participants, is described. The evaluation consists of three types of tests: complementary tests, preliminary tests, and performance tests. The complementary tests provide summary information about the source apportionment results as a whole. The preliminary tests check whether source/factors belong to a given source category. Three types of indicators: Pearson correlation (Pearson), standardized identity distance (SID), and weighted difference (WD) are used to test factor/ source chemical profiles, while factor/source time series and contribution-to-species values are tested only using the Pearson. The performance tests, based on international standards for proficiency testing, are targeted at evaluating whether the reported biases in the quantification of the factor/source contribution estimates (SCEs) and uncertainties are consistent with previously established quality standards in a fitness-for-purpose approach. Moreover, the consistency of the SCE time series is evaluated using a variant of the RMSE normalised by the reference standard uncertainty. The described methodology facilitates a thorough evaluation of the source apportionment output. The new indicator to compare source or factor profiles presented in this study (SID) is more robust and provides additional information compared to the existing ones. © 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license

Hereafter we present a laboratory evaluation of commercially available amperometric sensors for the monitoring of O 3 and NO 2 in ambient air at ppb level. The tests includes the determination of the response time of sensors, their calibration function, the evaluations of repeatability, short and long term drifts, the hysteresis effect and the matrix effect. Interferences from gaseous compounds, temperature and humidity parameters are also evaluated.

The KEY-VOCs project is a EURAMET joint research project focused on key Volatile Organic Compounds (VOCs) in air. One of its activities is the evaluation of sensors-based measurement systems. In Europe, the monitoring of benzene in ambient air is mandatory as set by the European Directive for air quality (AQD) [1]. This Directive states that the reference method of measurement shall consist of active or on-line sampling followed by gas chromatography [2]. These methods are time consuming, expensive to implement and not easily portable prohibiting more local estimation of the population exposure. However, the AQD allows using indicative measurements with higher uncertainty than those of the reference methods. Sensor systems are good candidates for indicative methods with the additional ability of near-to real-time measurements.

This article presents a literature review of sensors for the monitoring of benzene in ambient air and other volatile organic compounds. Combined with information provided by stakeholders, manufacturers and literature, the review considers commercially available sensors, including PID-based sensors, semiconductor (resistive gas sensors) and portable on-line measuring devices as for example sensor arrays. The bibliographic collection includes the following topics: sensor description, field of application at fixed sites, indoor and ambient air monitoring, range of concentration levels and limit of detection in air, model descriptions of the phenomena involved in the sensor detection process, gaseous interference selectivity of sensors in complex VOC matrix, validation data in lab experiments and under field conditions.

A growing number of companies have started commercializing low-cost sensors (LCS) that are said to be able to monitor air pollution in outdoor air. The benefit of the use of LCS is the increased spatial coverage when monitoring air quality in cities and remote locations. Today, there are hundreds of LCS commercially available on the market with costs ranging from several hundred to several thousand euro. At the same time, the scientific literature currently reports independent evaluation of the performance of LCS against reference measurements for about 110 LCS. These studies report that LCS are unstable and often affected by atmospheric conditions—cross-sensitivities from interfering compounds that may change LCS performance depending on site location. In this work, quantitative data regarding the performance of LCS against reference measurement are presented. This information was gathered from published reports and relevant testing laboratories. Other information was drawn from peer-reviewed journals that tested different types of LCS in research studies. Relevant metrics about the comparison of LCS systems against reference systems highlighted the most cost-effective LCS that could be used to monitor air quality pollutants with a good level of agreement represented by a coefficient of determination R2 > 0.75 and slope close to 1.0. This review highlights the possibility to have versatile LCS able to operate with multiple pollutants and preferably with transparent LCS data treatment.

The PM mass loss due to EDXRF analysis of PM 10 loaded filters was investigated in this study. The mass loss was gravimetrically evaluated on collocated filters analyzed both by EDXRF under vacuum condition and by Ion Chromatography. Then, this comparison was repeated with a gaseous mixture of air and Helium (He) instead of evacuating the measurements chamber of the EDXRF spectrometer. It was shown that EDXRF analysis under vacuum medium leads to mass loss of water soluble ions and other volatile constituents of PM 10 . The total PM mass loss was found not to be limited to water soluble ionic species, indicating that other volatile components of PM were likely to be lost. The use of a He and air mixture medium considerably reduced the PM mass losses. As the conclusion of the first step of this study, the mass loss of PM and ions mainly resulted from vacuum medium and to a lower extent to the level of x-Ray energy. In the second part of the study, the influence of EDXRF analysis using vacuum medium on the mass losses of ions were individually investigated. It was revealed that different amounts of NO 3 , NH 4 , Cl, Na, Mg and Ca were lost whereas SO 4 and K remained unchanged. At the last step of the study, the relationship between vacuum and application time was studied for quartz and Teflon filters. The longer the application time, the higher the PM mass loss was observed. It was also revealed that the behavior of mass loss of PM varies depending of filter materials. The mass loss of PM on quartz filter was shown by experiments to be 2 times higher than on Teflon filters., JRC.H.2-Air and Climate

We propose a geostatistical technique based on point-centred semi-variograms that can be used to derive valuable information about the spatial representativeness of air quality monitoring sites. Whereas classical geostatistics describes the spatial correlation structure of a concentration field in terms of the variogram, point-centred variography is based on the average of squared concentration differences observed in pairs formed from a particular central point and the set of all other points in the domain. It thereby places a monitoring station in the context of the local or regional air quality pattern. We demonstrate how a mathematical inversion of the point-centred variogram can provide estimates of the extent of the representativeness area of a monitoring site. The application of this approach is tested on a set of modelling data from the city of Antwerp, which was used for the FAIRMODE/AQUILA intercomparison exercise of methods for the assessment of spatial representativeness.

A prototype of an automatic static volumetric method based on the injection of a pure compound by means of a loop instead of a manual syringe was constructed, fitting for both laboratory use and field applications. The crucial point, the evaluation of the loop volumes, was solved developing two techniques allowing traceability to SI units. Gas mixtures of nitrogen monoxide (NO) in nitrogen in the 10−9 V/V range and carbon monoxide (CO) in air in the 10−6 V/V range were generated and successfully compared to other primary calibration techniques. Relative expanded uncertainties (k=2) of 1.1% (compared to 0.8% of the manual static volumetric method [1]) for NO gas mixtures and 0.3% CO gas mixtures were estimated.

The potential of the voltammetry method was examined for the determination of heavy metals in ambient air particulate matter (PM 10 ) on quartz filter. Cd, Pb, Cu, Zn, As were determined by anodic stripping voltammetry while adsorptive stripping voltammetry was used for the analysis of Ni. The method detection limit of these metals were 9.3, 0.1, 0.8, 0.3, 0.4, 0.1 ng m −3 for Zn, Cd, Pb, Cu, Ni and As, respectively. In addition, the analysis of a Certified Reference Material NIST 1648, yielded recoveries between 92% and 103%. Consequently, both the detection limit and recovery of the voltammetric method satisfy the requirements of the European Standard for the analyses of heavy metals in PM 10 (EN 14902). A comparison of the inductively coupled plasma-mass spectrometry (ICP-MS) and voltammetry method on the NIST 1648 and PM 10 filters showed the differences between them remained well within the level of uncertainty on the NIST 1648 requested by European Directives for heavy metals (25% for Pb and 40% for As, Cd and Ni, respectively). In addition to its compliance with legislations, the voltammetry method benefits from low investment cost and the potential of complete automation. As such, one may expect voltammetry to provide a reliable alternative to the European laboratories in charge of ambient air monitoring at the time when the European Directives require to measure heavy metals in PM 10 on a regularly basis.

This paper presents the results of a feasability study for the preparation and certification of reference materials (RMs) for nitrogen dioxide (NO(2)) and sulfur dioxide (SO(2)) in diffusive samplers. RMs for NO(2) were prepared by exposure to gas mixtures in a chamber while the RMs for SO(2) were prepared by liquid spiking. Certification of RMs for NO(2) was found feasible with a certified uncertainty of 5.8% and a proposed shelf life of 5 years. The uncertainty was calculated with contribution from the homogeneity of preparation, stability and transport of the CRMs and from an external verification of the certified value. To reach 5.8% of uncertainty, the contribution of the differences between the results of analysis by ion chromatography and colorimetry must be eliminated. It is proposed to solve this by pre-extracting the samplers with water before analysis. The results of this study indicate that the samplers are stable for at least two years before and after exposure when stored in a refrigerator. By contrast, the certification of RMs for SO(2) was found to not be feasible due to instability problems. This instability was attributed to reaction of sulfate on the walls of the samplers. Alternatively, the preparation of RMs by simultaneous exposure to SO(2) and NO(2) has been tested. Satisfying homogeneities has been reached both for NO(2) and SO(2).

Within the frame of the EU sponsored project ‘Cermatair’ (Certified Reference Materials for the Measurement of Gaseous Pollutants in Ambient Air, contract G6RD-CT-2001-00517) the feasibility of preparing and certifying reference materials for the measurement of sulphur dioxide, nitrogen dioxide, carbon monoxide and benzene in ambient air was studied. The project focused on measurements at concentration levels corresponding to limit values given in EU Ambient Air Quality Directives and covered reference materials for the reference methods specified in these Directives and for alternative methods based on diffusive sampling. State-of-the-art technologies for the production and certification of the reference materials were identified through literature surveys. Limited batches of reference materials were prepared and, wherever appropriate, tested for homogeneity. The reference materials were subsequently tested in small-scale external verifications, performed by 2–4 laboratories other than the preparation laboratory, aimed at identifying possible discrepancies between concentration values from the preparation processes and experimental values. The results of these verifications revealed possible mechanisms of certification (based on preparation or measurements). The remaining materials were subjected to a one-year stability study.

Hereafter, a modification of the traditional Palmes diffusion tube by fitting a membrane at the open end of the tube is presented. The membrane is placed to avoid the effect of turbulence due to wind speed. Further optimisations of the volume of absorbent and number of meshes retaining the absorbent are also described. Finally, the sampler is placed in an air proof container and is ready to use in order to improve the quality of measurement. In order to evaluate the performance of the membrane-closed Palmes diffusion tube to monitor nitrogen dioxide (NO2), experiments were carried out in an exposure chamber. The relative effect of NO2, temperature, humidity, averaging time and wind speed on the uptake rate of the sampler were then studied. Among the uncontrolled variables, temperature and humidity are found to be the most important ones leading to variability of the uptake rate. Two equations, used to estimate the uptake rate are established. These equations were then applied on some field data and laboratory inter-comparison tests. The agreement between the chemiluminescence method and diffusive samplers changes from 20% of mean bias with the traditional Palmes diffusion tubes to 5% of mean bias with the membrane-closed Palmes tube (MCPT) at the same monitoring site. In the laboratory inter-comparison, the MCPT has shown to be able to correctly measure NO2 under two extreme exposure conditions in contrast with other diffusive samplers. The uncertainty of the sampler is evaluated. It fulfils the European Directive data quality objective of 25% for accuracy and therefore the sampler qualifies as an indicative method for NO2 monitoring.

Using a clustered system of sensors able to measure NO/NO2, CO, CO2 and O3, we want to compare the performance of several calibration methods. Based on the outcome of the single sensor’s evaluations, the suitability of this cluster is investigated using deterministic or statistical methods such as multivariate regression and neural networks. The objective of study is to assess if the measurement uncertainty of sensors can reach the Data Quality Objective of the European Air Quality Directive for indicative methods., JRC.H.2-Air and Climate

This experimental study focusses on air quality monitoring by low-cost and accurate sensors to provide a rank of air quality indices for citizens community in smart cities. Inter-comparison long-term measurements of CO, NO 2 , SO 2 and PM 10 are performed in a real scenario using referenced chemical analyzers at air-quality monitoring stations. The accuracy of the low-cost sensors is assessed in order to address the Data Quality Objectives (DQO) of the EU Air Quality Directive 2008/50/EC for indicative measurements. These preliminary results show that the low-cost sensors, when accurate, are suitable to define a rank of individual air quality index (AQI) to inform effectively general public and enhance environmental awareness. Finally, we show that the developed multiparametric sensor-system NASUS can be integrated both in a distributed city-network based on cost-effective fixed nodes, and in portable handheld sensor-systems to monitor air-pollution personal exposure., JRC.H.2-Air and Climate

In this work we present a novel calibration procedure for resistive commercial sensors that consists on modeling the ozone sensor responses in presence of ambient humidity and temperature extracting its parameters and behavior and then combine it with a standard Partial Least Squares (PLS) regression. To assure the validity of the method then we apply it to the measurements of 19 different small commercial sensors from 5 different commercial brands in different conditions of ozone concentration, humidity and temperature. We compare the method with a regression without temperature and humidity and with a PLS regression in which we have taken into account the temperature and humidity but not modeled it. We see that this procedure has a mean decrease in the Root Mean Square Error. of 19.8%, JRC.H.2-Air and Climate

Monitoring personal exposure to particle matter (PM2.5) in ambient air requires performing measurements using portable monitors. In this work, the portable nephelometer SidePak AM510 Personal Aerosol Monitor manufactured by TSI Inc. was evaluated against a Tapered Element Oscillating Microbalance (TEOM) equipped with a Filter Dynamics Measurements System (FDMS). Conventionally, the SidePak is calibrated with respect to the Arizona Road Test Dust and then multiplied by an environmental calibration factor to yield mass concentration. To adapt this calibration to specific field conditions, we present an implementation of this calibration by introducing a growing factor correction which takes into account relative humidity and the dry and wet portions of the refractive index estimated from TEOM-FDMS measurements. PM2.5 sampling with several SidePaks AM510 was carried out in background and rural sites in the Po Valley (Italy). Modeled SidePak data were plotted vs. reference TEOM-FDMS data which show a good agreement., JRC.H.2-Air and Climate

Energy Dispersive X-ray Fluorescence (EDXRF) was compared to Inductively Coupled Plasma Mass Spectrometer (ICP-MS) for the measurements of elements (Mg, Al, Si, S, Cl, K, Ca, Ti, V, Cr, Fe, Co, Ni, Mn, Cu, Zn, As, Br, Sr, Pb, Mo, Cd, Sn and Sb) in particulate matter (PM10) collected on Teflon and two types of quartz filters at different sites. Two different calibration methods of EDXRF, linear calibration and standardless analysis, were studied. For the linear calibration, Pb, Mn, Fe, Cu and Zn were found to be site and filter type independent whereas Ca was only site independent. The site effect was evidenced for K, As, Ni, Ti, and V for quartz filter. The standardless EDXRF analysis showed better results than linear calibrations except for As, Co and V for Teflon filters and Cr and V for quartz filters. The estimated measurement uncertainties of standardless EDXRF analysis were calculated. They were found to be equal to or less than those field ones at certain masses for most of the studied elements. This suggests that the model equation to estimate measurement uncertainty of standardless EDXRF analysis covers all the uncertainty sources for certain elemental masses. It was found that standardless EDXRF analysis is able to quantify most of the elements studied, particularly on Teflon filters rather than quartz filters. The standardless EDXRF analysis complies with the data quality objectives (DQO) of European Directives to measure Pb in PM10 for three types of filters, even at concentrations lower than limit values (LV). The detection limits (MDL) of standardless EDXRF analysis for measuring As and Cd were found to be insufficient to meet the legislative requirements. The MDL of Ni was sufficiently low for measurements; however, measurement uncertainties remained higher than the DQO at the lower concentrations than LV., JRC.H.2-Air and Climate

This paper presents the results of an intercomparison exercise for the determination of arsenic (As), cadmium (Cd), nickel (Ni) and lead (Pb) in PM10, which are regulated by the European Directives for ambient air quality. Thirteen laboratories participated, generally using the European reference methods of measurement which consist of a microwave digestion followed by analysis with either ICP-MS or GFAAS. Each participant was asked to analyse five test samples: a liquid Certified Reference Material (CRM), two sub-samples of a NIST dust CRM (one already-digested and one to be digested by the participants) and two loaded filters (one already-digested and one to be digested by the participants). Participants were able to meet the Data Quality Objectives (DQOs) of the European Directives (expanded uncertainties of 25% for Pb and 40% for As, Cd and Ni) for 93% of all test samples, except for the loaded filter that were digested by the participants. In fact, only 76% of DQOs were met for this test sample, the closest to a routine sample analysis in the laboratory. The difficulties in analysing this test sample came mainly from digestion and contamination processes. Satisfactory results were also obtained using other digestion techniques (Soxhlet extraction and high pressure methods) and analytical methods (ICP-OES for Cd, Ni and Pb, EDXRF for Pb and Ni and Voltammetry for As, Ni, and Pb). Participants claimed uncertainties of about 10% for Pb and between 15 and 20% for As, Cd and Ni. These uncertainties were confirmed for 77% of results. The reproducibility of the methods of measurement was between 41 and 54% while repeatability remained between 5 and 12% except for the analysis of As on filters which was up to 20%. The majority of participant results showed higher between-day variability (14 ± 11%) than within-day variability (6.0 ± 5.3%).

A study of the equivalence to the reference methods of the Radiello samplers for ozone (O(3)) and benzene as well as the membrane-closed Palmes tube (MCPT) for nitrogen dioxide (NO(2)) is presented. These samplers benefit from new model equations capable of estimating their uptake rate. For O(3), the aim here was to demonstrate the equivalence for the reference period of 8 h and 120 microg m(-3), the target value of the 3rd European Daughter Directive. For benzene, the demonstration of equivalence to the annual limit value of 5 microg m(-3) of the 2nd European Daughter Directive was examined. In the case of NO(2), the equivalence to the annual limit value of the 1st European Daughter Directive (40 microg m(-3)) was considered. Results show that the radial sampler for O(3) fails to meet the Data Quality Objective (DQO) for continuous monitoring. However, with an expanded uncertainty of less than 30%, the O(3) diffusive sampler fulfils the DQO for indicative measurements. For benzene, the Radiello sampler exposed for 7 days gave satisfying results showing the ability of the sampler to meet the DQO of the reference method. Nevertheless, the field tests should be complemented by measurements for a wider range of benzene concentrations. In the case of NO(2), all the results of the laboratory and field experiments respected the requirements necessary for the demonstration of equivalence. Overall, these findings thus show that the Radiello sampler and the MCPT are equivalent to the reference methods only for assessment of benzene and NO(2), respectively.

The need to collect data representative of overall urban pollution is all-important in order to monitor the population exposure. High spatial resolution monitoring using diffusive samplers allows studying of the urban pollutant distribution, thus enabling deeper investigation of their generation and diffusion mechanisms. Nevertheless, such a monitoring campaign has a certain cost. In this study we point out how to find the best compromise between the number of necessary measurements and the affordable costs for monitoring campaigns. We also describe an innovative method for the proper design of a fixed urban monitoring network by means of preliminary high spatial resolution campaigns using diffusive samplers. Four European capital cities (Dublin, Madrid, Paris and Rome) were monitored six times, each time for seven days. Benzene, toluene, ethylbenzene, xylenes (BTEX) and NO(2) concentrations were measured at 146 sites in Dublin, 293 in Madrid, 339 in Paris and 290 in Rome. Multiscale grids have been drawn which ranged in mesh size from 500 m to 2 km. The statistical processing of data produced a twofold result: the creation of isoconcentration maps with geostatistical procedures, and an algorithm aimed at locating the minimum number of sampling sites where the fixed monitoring stations should be placed. Average urban levels estimated on the basis of these selected sites differ by less than 8% from those calculated on the whole populations of the sampled points. The aim of this work is to investigate how far the resolution of a monitoring campaign of urban pollution by diffusive sampling can be reduced, thus making the monitoring less expensive in terms of human and financial resources, while preserving the same quality of the results that could be achieved with a higher resolution. We found that there is no significant loss of information when the resolution of the monitoring grid for BTEX is lowered to a mesh size of 1.85 km, that is a sampling site each 3.4 km(2), and that the minimum number of sampling sites to be used is N = 0.29 A, where A is the urban surface to be monitored (in km(2)). As the spatial distribution of NO(2) is less sensitive to the distance from the emission source than that of BTEX, this relationship could be retained as a valid lower limit for the mesh grid size also for NO(2) monitoring.

The uncertainty of benzene measurements obtained by the analysis of thermally desorbable Radiello diffusive samplers was evaluated according to the recent standard EN 14662-4 [EN 14662-4, 2005. Ambient air quality. Standard method for measurement of benzene concentrations. Part 4: diffusive sampling followed by thermal desorption and gas chromatography]. Considering the results of laboratory experiments, all the sources of uncertainty regarding the diffusive sampler method characteristics were accessed for the sampling times of 7 and 14 days. The major part of the uncertainty budget (>79%) was explained by the variation of the sampling rate due to the environmental factors (temperature and concentration level). For weekly sampling, the diffusive sampler method satisfies the data quality objectives of the European Directive to supply the indicative measurements as well as the reference measurement, since the expanded uncertainty is found −3 , which is generally observed in environmental air monitoring. This narrow validation domain limits the application fields of the Radiello sampler exposed for 7 days to indoor air, personal exposure and ambient atmospheric monitoring excluding workplaces. For 2-week sampling, the expanded uncertainty of measurements exceeds 30%. However, this diffusive sampler can still be used to carry out an objective evaluation of benzene (minimum quality objective for the accuracy of 100%). Therefore, the performance of this diffusive sampler method appears to be suitable for the benzene monitoring in ambiant air.

The 8-h ozone radial diffusive sampler was evaluated according to the CEN protocol for the validation of diffusive samplers. All the parameters regarding the sampler characteristics were found to be consistent with the requirements of this protocol apart from the blank value, which must be evaluated and subtracted at each sampling. The nominal uptake rate was determined in laboratory conditions. However, the uptake rate depends on the mass uptake, temperature, humidity and on the combination of temperature and humidity. Based on laboratory experiments, an empirical model has been established which improved the agreement between the radial sampler and the reference method. This improvement was observed under several different meteorological and emission conditions of sampling. By using the model equation of uptake rate, the data quality objective of 30% for the expanded uncertainty included in the O 3 European Directive, is easily attained. Therefore, the sampler represents an appropriate indicative method.

This study presents an evaluation of the extent of differences between measurements performed by O(3) and NO(2) diffusive samplers and by the reference methods for diffusive samplers commercially available. The tests were performed in an exposure chamber under extreme conditions of controlling factors and under field conditions. For NO(2), the results of the laboratory experiments showed that most of the diffusive samplers were affected by extreme exposure conditions. The agreement between the samplers and the reference method was better for the field tests than for the laboratory ones. The estimate of the uptake rate for the exposure conditions using a model equation improved the agreement between the diffusive samplers and the reference methods. The agreement between O(3) measured by the diffusive samplers and by the reference method was satisfactory for 1-week exposure. For 8-hour exposures, the diffusive samplers with high uptake rates quantified better the O(3) concentration than the samplers with low uptake rates. As for NO(2), the results of the O(3) field tests were in better agreement with the reference method than the ones of the laboratory tests. The field tests showed that the majority of diffusive samplers fulfils the 25% uncertainty requirement of the NO(2) European Directive and the 30% uncertainty requirement of the O(3) European Directive for 1-week exposure.

The influence of environmental parameters on the uptake rate of the nitrogen dioxide (NO(2)) Palmes diffusive sampler was investigated. The main factors affecting the uptake rate were found to be wind speed, the preponderant factor, followed by relative humidity and temperature. The NO(2) concentration and exposure time, as well as the interactions among the factors were not found to have a significant influence on the uptake rate of the Palmes sampler. As a result, a model able to predict the uptake rate of the Palmes sampler was established. In addition, by using the model-predicted uptake rate, the agreement between chemiluminescence and the Palmes sampler during the field tests was improved. The NO(2) Palmes diffusive sampler was shown to comply with the requirement on accuracy defined by the European Directive for the indicative methods of measurements.

Cette etude presente un essai d’aptitude pour les Associations agreees de surveillance de la qualite de l’air effectuant la mesure du NO2 par echantillonneur a diffusion en France. L’evaluation des donnees montre que les participants obtiennent de bons resultats pour les expositions sur sites. Pour les essais sur tubes dopes et materiaux de reference, environ la moitie des participants obtiennent de bons resultats alors que l’objectif de qualite etait aussi plus ambitieux. L’intercomparaison a egalement ete planifiee pour determiner la repetabilite et la reproductibilite de la methode de mesure. Pour les essais sur site, la repetabilite diminue avec le niveau de concentration mesuree, passant d’environ 20 a 10 % quand les concentrations evoluent de 15 a 40 μg/m3. En revanche, la reproductibilite semble stable a 45 %. La repetabilite (10 %) et la reproductibilite (20 %) obtenues pour les essais sur les tubes dopes et les materiaux de reference sont inferieures aux repetabilite et reproductibilite des essais sur site.La reproductibilite est toujours superieure a la repetabilite, ce qui suggere la presence d’erreurs systematiques dans la preparation et/ou l’analyse des echantillonneurs. De plus, le controle qualite efffectue sur la methode d’analyse montre que la valeur du blanc, l’ecart de linearite et la variation de la pente d’etalonnage peuvent etre diminues.Globalement, les resultats de cette etude apparaissent plus precis que ceux qui sont obtenus dans le cadre du programme de controle qualite WASP et du reseau national NO2 du Royaume-Uni. Cette etude montre egalement une amelioration des resultats par rapport a l’intercomparaison 1999 realisee par l’Ecole des mines de Douai.Le resultat le plus notable reste que la mediane des valeurs moyennes de tous les laboratoires pour les essais sur site est en accord parfait avec les mesures des analyseurs par chimiluminescence. De plus, les nouveaux materiaux de reference prepares par l’ERLAP en exposition sous atmosphere controlee, ont demontre leur validite pour l’utilisation dans ce type d’intercomparaison

Hereafter, an assessment of the ability of the chemiluminescence method to measure ambient NO2 with an accuracy within 15%, as requested by the data quality objective of European directive 1999/30/CE, is presented. In general, uncertainty is evaluated using the response to reference materials or by means of inter-comparisons used to determine some statistics like repeatability, reproducibility and calibration bias. These are incomplete approaches and the method of the Guide to the Expression of Uncertainty in Measurement, advised by the Directive, should be preferred. In fact, even if it requires a large data set, it allows the relative influence of all possible sources of uncertainty to be studied. The extent of NO2 uncertainty is mainly dependent on the level of NO. It is decreased by NOx and the correlation between NOx and NO. Furthermore, the uncertainty budget reveals that the contribution of accuracy of calibration standard, linearity, converter efficiency and drift of the analyser between calibration checks to the overall uncertainty is less important than the contribution of interference, mainly humidity and PAN in rural areas. The relative expanded uncertainty of the NO2 hourly average exceeds 30% for NO2 concentrations lower than 40 microg m(-3). Nevertheless, the data quality objective of 15% is reached for 200 microg m(-3), the hourly limit value of the European directive. On the contrary, at the limit value on the annual average, 40 microg m(-3), the data quality objective is not met if NO is higher than 100 microg m(-3). However, the data quality objective could be reached by correcting the measurements with the bias due to interference.

The European Reference Laboratory of Air Pollution implements the static volumetric method for the preparation of nitrogen monoxide and sulphur dioxide reference standard gas mixtures. According to the new ISO guide for the expression of uncertainty, the uncertainty of these standards is up to 0.8% for nitrogen monoxide in the range 100 to 600 ppbv, and up to 0.4% for sulphur dioxide in the range 200 to 400 ppbv. The values presented in the present paper suggest that there is a 95% probability of the true value lying within the interval specified. To attain such low uncertainty values, the standard procedure for the implementation of the static volumetric method must be rigorously followed, and instruments must be carefully maintained.

Systematic collection of long term meso- to large-scale datasets of ambient air quality provides an indispensible means for air pollution monitoring. However, the quality of these monitoring data depends on the chosen method of measurements and the QA/QC procedures applied. We present the first version of a prototyped screening tool for the automatic detection of outliers in large data volume air quality monitoring records. The method is based on an adaption of the existing Smooth Spatial Attribute Method, which considers both attribute values and spatio-temporal relationships. An application example of the method is demonstrated by computing warnings on abnormal records in the 2006/2007 time series of PM 10 daily values of background stations reported in the European air quality database AirBase.

The first key comparison on carbon monoxide (CO) in nitrogen dates back to 1992 (CCQM-K1a). It was one of the first types of gas mixtures that were used in an international key comparison. Since then, numerous national metrology institutes (NMIs) have been setting up facilities for gas analysis, and have developed claims for their Calibration and Measurement Capabilities (CMCs) for these mixtures. Furthermore, in the April 2005 meeting of the CCQM (Consultative Committee for Amount of Substance) Gas Analysis Working Group, a policy was proposed to repeat key comparisons for stable mixtures every 10 years. This comparison was performed in line with the policy proposal and provided an opportunity for NMIs that could not participate in the previous comparison. NMISA from South Africa acted as the pilot laboratory. Of the 25 participating laboratories, 19 (76%) showed satisfactory degrees of equivalence to the gravimetric reference value. The results show that the CO concentration is not influenced by the measurement method used, and from this it may be concluded that the pure CO, used to prepare the gas mixtures, was not 13C-isotope depleted. This was confirmed by the isotope ratio analysis carried out by KRISS on a 1% mixture of CO in nitrogen, obtained from the NMISA. There is no indication of positive or negative bias in the gravimetric reference value, as the results from the different laboratories are evenly distributed on both sides of the key comparison reference value., JRC.H.2-Air and Climate