the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 21 Jan 2022
Evaluating uncertainty in sensor networks for urban air pollution insights
Ambient air pollution poses a major global public healthrisk. Lower-cost air quality sensors (LCSs) are increasingly being exploredas a tool to understand local air pollution problems and develop effectivesolutions. A barrier to LCS adoption is potentially larger measurementuncertainty compared to reference measurement technology. The technicalperformance of various LCSs has been tested in laboratory and fieldenvironments, and a growing body of literature on uses of LCSs primarily focuses onproof-of-concept deployments. However, few studies have demonstrated theimplications of LCS measurement uncertainties on a sensor network's abilityto assess spatiotemporal patterns of local air pollution. Here, we presentresults from a 2-year deployment of 100 stationary electrochemical nitrogendioxide (NO2) LCSs across Greater London as part of the Breathe London pilot project (BL). We evaluated sensor performance using collocations withreference instruments, estimating∼ 35 % average uncertainty(root mean square error) in the calibrated LCSs, and identified infrequent,multi-week periods of poorer performance and high bias during summer months.We analyzed BL data to generate insights about London's air pollution,including long-term concentration trends, diurnal and day-of-week patterns,and profiles of elevated concentrations during regional pollution episodes.These findings were validated against measurements from an extensivereference network, demonstrating the BL network's ability to generate robustinformation about London's air pollution. In cases where the BL network didnot effectively capture features that the reference network measured,ongoing collocations of representative sensors often provided evidence ofirregularities in sensor performance, demonstrating how, in the absence ofan extensive reference network, project-long collocations could enablecharacterization and mitigation of network-wide sensor uncertainties. Theconclusions are restricted to the specific sensors used for this study, butthe results give direction to LCS users by demonstrating the kinds of airpollution insights possible from LCS networks and provide a blueprint forfuture LCS projects to manage and evaluate uncertainties when collecting,analyzing, and interpreting data.
- Article
(3996 KB) - Full-text XML
- Supplement
(2081 KB) - BibTeX
- EndNote
Ambient (outdoor) air pollution is a leading contributor to human diseaseand mortality around the world, causing more than 4 million prematuredeaths annually, with the greatest health burden in low- and middle-incomecountries (WHO, 2018; HEI, 2020). Within cities, the burden of air pollutionis not distributed equally, with significant spatial heterogeneity insources, concentrations, and exposures (e.g., Apte et al., 2017; Clark etal., 2014; Miller et al., 2020; Shah et al., 2020). Many of the world's mostpopulous and polluted regions are also those with limited air qualitymonitoring infrastructure, restricting the potential for data-driven airquality management or public awareness campaigns (Pinder et al., 2019). Evenin many high-income countries, ambient air pollution monitoring isrelatively sparse (e.g., Apte et al., 2017; US GAO, 2020). Referencemonitoring stations are state of the art in terms of accuracy andreliability and are required for regulatory reporting (EU, 2008). However,they are costly (∼ 104–105 USD).
Lower-cost air quality sensors (LCSs) are increasingly being explored as analternative or supplement to reference monitors. LCSs are orders of magnitudeless expensive (∼ 102–104 USD) and are thereforemore suitable for dense deployments. They are commercially available fromnumerous manufacturers, and the market is expanding rapidly. The literatureon LCSs has primarily focused on technical evaluations of sensor performancein laboratory or field settings (Castell et al., 2017; Duvall et al., 2016;Jiao et al., 2016; Karagulian et al., 2019; Kelly et al., 2017; Lewis etal., 2016; Mead et al., 2013). Comprehensive reviews of sensor technologyhave identified common performance issues including drifting baselines andcross-interference from other pollutants as well as sensitivity toenvironmental conditions such as temperature and relative humidity (WMO,2021). The literature also presents a variety of approaches for improvingthe accuracy of unprocessed sensor data, including calibrations usingcollocations with reference instruments, in-field calibrations withoutcollocations, and machine learning techniques, among others (Kim et al.,2018; Munir et al., 2019; Sahu et al., 2021; Spinelle et al., 2015;Zimmerman et al., 2018).
A growing body of literature on uses of LCSs primarily focuses on scientificapplications and proof-of-concept deployments. Case studies havedemonstrated the potential for LCS networks to provide data insights about alocal air pollution environment, including characterizing spatiotemporaltrends in ambient air quality (Castell et al., 2018; Caubel et al., 2019;Mead et al., 2013; Pope et al., 2018; Popoola et al., 2018) and improvingair quality models through data fusion or assimilation (Bi et al., 2020;Carruthers et al., 2019; Gupta et al., 2018; Lopez-Restrepo et al., 2021).While previous LCS deployments often consider uncertainty of individualsensors relative to a reference instrument, we are unaware of networkdeployments where the spatiotemporal observations have been directlycompared to results from a reference network.
As LCS technology becomes more ubiquitous, there is growing interest fromgovernments and civil society to use data from LCS monitoring networks inair quality assessment and urban planning. To manage the inherentuncertainties in LCSs, guidance is needed on how users can evaluate sensorperformance and decide on the most appropriate and robust uses of theirdata. In this work, we evaluate a sensor network's ability to characterizespatiotemporal air pollution patterns in the megacity of Greater London byusing data from an LCS monitoring network deployed as part of the BreatheLondon pilot project (BL).
London was an ideal study area for LCS evaluation due to the city'sextensive network of reference air pollution monitors as well as a range ofadditional tools, including a detailed emissions inventory andhigh-resolution modeling, all of which contribute to an advancedunderstanding of historical and current air pollution (GLA, 2021). Further,while air pollution has improved in recent years, in 2019 an estimated 3600to 4100 premature deaths were attributable to anthropogenic fine particulatematter (PM2.5) and nitrogen dioxide (NO2) in London alone (Dajnaket al., 2021), and pollutant concentrations remain above UK and WHO guidelinelevels in many areas of the city (GLA, 2020a). In 2021, the Court of Justiceof the European Union ruled that the UK has been exceeding legal limits ofNO2 since 2010 and that the government failed against its legal dutiesto put timely mitigation plans in place (The Guardian, 2021). This workfocuses on NO2 data, which were a key measurand of the project based onthe local regulatory priorities.
We first evaluated the performance of a subset of NO2 sensors thatwere collocated with reference instruments. The uncertainties determinedfrom these evaluations were then considered in the context of specificanalysis applications, or “use cases”, of LCS data, including long-termconcentration trends, temporal concentration patterns (i.e., diurnal andday-of-week), and quantification of regional episodes of elevated airpollution. LCS network results were compared to results from an extensivenetwork of London reference monitors, demonstrating the extent to which theBL network produced accurate spatiotemporal insights about air pollution and illuminating how sensor uncertainties identified during collocationsaffected the network's ability to characterize local air pollution.
While the BL LCS results show many areas of agreement with reference networkdata, with some areas of discrepancy, the comparisons are onlyrepresentative of a selected sensor technology (electrochemical NO2sensors of a specific vintage from a specific supplier) deployed in aspecific environment type; care should be taken in extrapolating results toother sensors and environments (i.e., differing pollution levels and weatherconditions). Nevertheless, the methods and lessons presented here can aidthe design and operation of future LCS deployments by providing a blueprintfor users to quantify and manage uncertainty in their own LCS datasets andexplicitly consider the implications when investigating locally relevant airpollution questions.
2.1 Monitoring devices
The BL NO2 dataset includes data from 100 AQMesh units (EnvironmentalInstruments Ltd., Firmware V 3.24), commercially available devices whichhave been previously tested and utilized by researchers and air qualitymanagers (Fig. 1b) (AQMesh, 2021; AQ-SPEC, 2015; Castell et al., 2017). Adetailed description of the AQMesh units can be found elsewhere (e.g.,Castell et al., 2017). AQMesh measurements of nitrogen dioxide (NO2),the focus of this paper, relied on an Alphasense Ltd. O3-filteredelectrochemical sensor. The AQMesh devices in BL also measured nitric oxide(NO), particulates (PM2.5 and PM10), and carbon dioxide (CO2),and 10 devices additionally measured ozone (O3).
Figure 1(a) BL network locations across Greater London.(b) Picture of BLAQMesh unit (indicated by arrow) installed at Kew Road, Richmond.
2.2 Network design and deployment
We deployed AQMesh units across Greater London (Fig. 1) in areas identifiedin consultation with the Greater London Authority (GLA), though finallocations depended on obtaining permissions from site owners. We soughtlocations across a range of traffic levels and at varying distances frommajor roads and intersections, parks, residential areas, high-trafficstreets, and other commercial areas. In addition, we included monitoring atsensitive receptors, including some primary schools and medical facilities.
Each BL location was classified by site type (kerbside, roadside, or urbanbackground) based on the local characteristics in accordance with GLAguidance for London air quality monitoring (GLA, 2018). Kerbside locationswere usually within 1 m of a road and were expected to have highpollutant concentrations where traffic was the dominant source. Roadsidelocations were also situated near roads (usually<5 m) but wereexpected to be more representative of pedestrian exposure. Urban backgroundlocations were mostly sited within school yards away from dominant emissionssources such as busy roads. BL AQMesh devices were often installedmarginally higher (∼ 3–4 m) than London reference monitors(∼ 2 m) to avoid physical tampering. Some monitors that werewithin 1 m of the road were still classified as urban background orroadside based on judgment of local features, including device height,positioning, and proximity to sources. While prevailing guidance recommendsdevices be placed away from structures, with270∘ unobstructed flow,this goal was not achieved at many sites where the only option forinstallation and power supply was on a building façade (EU, 2008). Thus,classifications are informative but somewhat imperfect.
Of the 112 AQMesh locations in the NO2 dataset (number exceeds 100because some sensors were relocated during the project), 36 sites wereclassified as kerbside, 36 as roadside, and 40 as urban background. Thelocations and site types are shown in Fig. 1a.
2.3 Data collection, processing, and QA/QC
We evaluated AQMesh measurements of NO2 collected during the BreatheLondon pilot project from September 2018 through November 2020 (BreatheLondon, 2021a). The devices were set to take a measurement every 10 sand delivered averaged readings every minute (i.e., an average of sixreadings). These 1 min data were transferred using a built-in GPRS modemto the manufacturer's (Environmental Instruments Ltd.) server in nearreal time, where they were processed by the manufacturer using proprietaryalgorithms based on their factory testing, and are termed here as prescaleddata. Individual data points were accompanied by flags regarding sensorstatus. Data were then ingested into a data platform hosted by ACOEM AirMonitors Ltd., who also managed the monitor deployment, maintenance, andmanual QA/QC process (Breathe London, 2020). Cambridge EnvironmentalResearch Consultants (CERC) applied a sensor-specific calibration gain andoffset (see Sect. 2.3.1) to each device's 1 min prescaled data to producea calibrated dataset. CERC then filtered data for valid flags and high andlow limits that screened out physically unrealistic concentrationmeasurements and averaged measurements to hourly time resolution using an85 % data capture threshold per hour. Manual inspection of sensor data wasperformed weekly to identify anomalous measurements. If a sensor malfunctionwas identified through QA/QC protocols, ACOEM Air Monitors Ltd. techniciansintervened to mitigate the issue, usually through replacement of faultysensors.
2.3.1 Sensor calibration
NO2 sensors in the field (termed “candidate sensors” here) werecalibrated using one of three methods: reference site collocation, transferstandard collocation, or remote network calibration method. For referencesite collocations, a candidate AQMesh unit was installed alongside areference monitor from the London Air Quality Network (LAQN) or UK AutomaticUrban and Rural Network (AURN) (Fig. S1 shows a picture of an examplereference site collocation). Transfer standard collocations relied on nineAQMesh devices that were periodically (every 2–4 months) collocated andcalibrated against reference monitors; these calibrated AQMesh units werethen used as transfer standards and were collocated with so-called candidateAQMesh units in the field to determine the latter's calibration parameters.The duration of typical calibration collocations was 7–14 d (for bothreference and transfer standard methods), though long-term collocations werealso conducted for further performance evaluation purposes. Calibration gainand offset parameters were obtained by performing a linear regression on thehourly averaged collocation time series after excluding the 1st and99th percentile of hours during the collocation based on the ratio ofreference candidate values. Calibration parameters were deemed valid andapplied to the candidate sensor if the scaled collocation time series metstatistical criteria of normalized root mean square error (nRMSE)<50 % (Eq. 3) andR2>0.7 (Eq. 4), which ensured that sensor performance wassufficient to calculate robust calibration parameters and effectivelyexcluded periods where the NO2 variability was too low to provide ameaningful test of sensor gain and offset.
The remote network calibration method is a novel approach,developed and applied by the University of Cambridge project team, thatremotely derives unit-by-unit calibration parameters for the entire sensornetwork in lieu of physical collocations. The algorithm uses a spatial scaleseparation methodology described in previous work (Heimann et al., 2015;Popoola et al., 2018) to calibrate sensors in relation to each other whenpollution levels are consistent across the network and obtains traceability(connection to reference standard with known uncertainties) from a singlecalibrated reference monitor (Popoola et al., 2022; Popoola and Jones, 2020). For BL, a single (site-dependent) calibration was performedusing the period May–December 2019 and applied to the entire dataset. Thispaper does not intend to evaluate the network calibration methodcompared to other approaches. The method and its performance will beaddressed in more detail in a separate study (Popoola et al., 2022). However, we describe the method here because it was used toscale a subset of BL sensors which had no physical (reference or transfer)calibration available, and we include data from this subset of sensors tomaximize the number of sensor locations in our analysis and comparisons tothe reference network.
When multiple valid calibration options were available for a specific AQMeshsensor in the network, a decision tree was used which prioritized (i)reference site collocation (n=11), (ii) transfer standard collocation(n=73), and (iii) network calibration (n=38); the total number ofcalibrations applied exceeded the number of devices because failed sensorswere replaced and re-calibrated.
2.3.2 Ozone cross-interference correction
A long-term upward drift in BL NO2 sensor measurements was observed(Fig. S2), which we hypothesized to be caused by an ozonecross-interference. A correction was applied to the hourly NO2 datasetthat subtracted a fraction of the derived site-specific O3concentration from the scaled NO2 readings. Site-specific O3 wasdeduced using upwind background reference O3 measurements; underlow-NOx conditions (<10 ppb) the site-specific O3 was assumedto be the upwind background O3 concentration; otherwise it was assumedto be the difference between background O3 and the site-specific NOconcentration. Because the effect appeared to increase as a sensor aged, thecross-sensitivity correction for ozone was assumed to start at 0 % uponinitial sensor deployment and exponentially increase to a maximum of+18 % of estimated site-specific ozone concentrations 6 months later.Figure S2 illustrates the effect of the correction on BL network meanNO2 concentrations throughout the campaign. Figures S3 and S4 showevidence supporting the ozone cross-interference hypothesis and anevaluation of the correction method for an individual sensor. Note that dueto a complex set of factors including the combination of factory (AQMesh)and field calibration methods, we could not exclude other possible causes ofobserved irregularities in sensor measurements. Except for the short-termcollocation analysis results (Fig. 2), the results presented throughout thispaper use the scaled hourly average ozone-corrected dataset.
Figure 2Performance of calibrated sensors during short-term (typically 7–14 d) collocations with reference instruments. Unfilled circles arecollocations that started in July 2019 during periods of elevatedtemperatures. Statistics calculated from hourly measurements (Eqs. 1–4).
Detailed documentation of the static network QA/QC procedures is availablein the project QA/QC manual in the Breathe London Technical Report (BreatheLondon, 2020).
2.4 Reference and meteorological data
Hourly NO2 and O3 concentration data were downloaded for 105reference monitors within Greater London that were classified as kerbside(n=12), roadside (n=62), or urban background (n=31) using the Ropenair package (Carslaw and Ropkins, 2012). These monitors, which we referto collectively as the “reference” network, include sites from multipleoverlapping UK networks including the London Air Quality Network (LAQN), AirQuality England (AQE) network, and Automatic Urban and Rural Network (AURN).At the time of download (9 June 2021) reference data were fully ratified for69 sites. At 36 sites, some 2020 data were categorized as provisional andare thus subject to change during the ratification process. Hourly ambientair temperature observations at London Heathrow Airport, located within theGreater London study region and∼ 25 km west of CentralLondon, were accessed from the National Oceanic and AtmosphericAdministration (NOAA) Integrated Surface Database (ISD) via the R worldmetpackage (NOAA, 2021; Carslaw, 2020).
2.5 Sensor performance statistics
The reference site collocations described in Sect. 2.3.1 were alsoused to evaluate sensor performance. A total of 98 collocations wereperformed between a lower-cost (LC) sensor and a reference monitor, including 10 sensorsthat were collocated more than once and 2 sensors that were collocated forlong-term periods of>80 weeks. The statistics in Eq. (1)–(4) wereused to evaluate sensor performance during reference site collocations (arepresentative example of collocation results is shown in Fig. S5). Thefollowing statistics were calculated from hourly time series data for eachindividual collocation ofn h duration:
where Sen represents the BL sensor measurement, and Ref represents the observedreference measurement.
2.6 ADMS-Urban modeling data
The ADMS-Urban air pollution dispersion model was used to simulate 2019hourly NO2 concentrations at BL and reference network monitoringlocations in order to estimate the expected difference in NO2 pollutionlevels between the two networks (McHugh et al., 1997). The model usedtraffic flows and speeds and 1 km gridded emissions of NO2 from theLondon Atmospheric Emissions Inventory (LAEI) 2013 dataset (published in2016), interpolated to 2019 from the 2013 base year and 2020 futurepredictions, combined with road traffic emissions factors from the EmissionFactor Toolkit (EFT) v8 for 2019 and real-world adjustment factors tocalculate road source emissions. The model includes atmospheric chemistry aswell as complex urban effects including street canyons and urban canopy.Individual monitoring sites were modeled as discrete receptors with theappropriate position and height. NO2 sources from outside the modeleddomain were represented using hourly background concentrations at one offour rural AURN (Automatic Rural and Urban Network) stations located outsideGreater London, based on which station was upwind at that hour, and hourlymeteorological data were used from London Heathrow Airport. The modelingscenario (“Hotspot 2019”) includes weekday diurnal emissions patterns torepresent variations in traffic flow and improvements to LAEI traffic flow.Additional details on the ADMS-Urban model and Hotspot 2019 scenario areavailable in the Breathe London Technical Report (CERC, 2021). To calculatethe modeled difference between BL and reference network means for the year2019 (Sect. 3.2.1), we selected all monitor hours with valid model–observation pairs(i.e., a valid modeled and observed concentration existed at that hour) forall reference and BL sites analyzed in this paper. The modeled 2019means were calculated for each network from the pooled monitor hours.
3.1 Network performance
3.1.1 Data capture
The BL network generated nearly 1.5 million hourly calibrated NO2measurements from 100 devices at 112 locations over the course of the26-month pilot campaign. The number of sensor locations producing valid,calibrated data gradually increased over the first 7 months (Fig. S6).The initial delay in network data capture was caused by logisticalchallenges faced at the outset of the project, including obtainingpermissions for monitor deployment and conducting calibrations for eachsensor. By the spring of 2019 the majority of the network was operationaland generated valid data for the remainder of the project, though the totalnumber of NO2 sensors producing valid data fluctuated due toredaction of flagged data and the downtime of sensors that failed during theproject before replacement and re-calibration were performed. In total, 35NO2 sensors were replaced due to failure, with most failures occurringduring the winter. Additional considerations and lessons learned forstationary sensor network setup and maintenance are discussed in the BreatheLondon Blueprint (Breathe London, 2021b).
3.1.2 Measurement uncertainty of calibrated sensors
Figures 2 and 3 present measurement uncertainty statistics for calibrated BLLCSs based on short-term (typically 7–14 d) and long-term (>80weeks) reference collocations. Both analyses quantify uncertainty of sensormeasurements that were calibrated based on results of a prior reference sitecollocation (Sect. 2.3.1). These results allow us to evaluate theeffectiveness of the project's QA/QC procedures (including calibration)since each repeat collocation serves as an independent test of theproject-long uncertainties in sensors that were calibrated during a discretetime period.
Figure 3Performance of two calibrated sensors during long-term referencecollocations. Sensors were calibrated using linear regression against thereference instrument during a 2-week collocation directly preceding theevaluation period (calibration period not shown).(a) Daily mean NO2concentration time series comparison of BL sensor and reference monitormeasurements.(b) MBE (Eq. 1) and RMSE (Eq. 2) statistics of hourly BLsensor measurements compared to reference measurements during 14 dperiods.(c) Scatterplot and statistics (Eqs. 1–4) comparing hourly BLsensor (x axis) and reference monitor (y axis) measurements for entireevaluation period.
Figure 2 shows evaluation results for 10 calibrated sensors that werecollocated for subsequent short-term periods (typically 7–14 d) thatbegan 1–84 weeks after an initial reference site calibration period. Thesesubsequent collocation periods (n=35) were used to estimate calibratedsensor uncertainty compared to reference measurements (e.g., unit 99 wascalibrated based on the first reference site collocation in October 2018;uncertainty statistics in Fig. 2 were calculated from the second and thirdreference collocations, which occurred in April and July of 2019).
A medianR2 of 0.79 indicates that calibrated sensors effectivelycaptured changes in NO2 concentrations that were measured by referenceinstruments. The median MBE was 8.0 µg m−3 (23 % of meanconcentration) with a range of−19 to 34 µg m−3 (−37 % to 121 %of mean concentration), and the median normalized RMSE was 35 % (range of16 % to 189 %). The large range of biases exhibited by individual sensorsand the systematically high median bias of the collocated sensors revealvariability in the consistency of sensor response over time (and underdifferent meteorological conditions) and serve to assess the robustness ofinitial sensor calibrations when applied to a longer time series. However, wenote that uncertainty statistics in Fig. 2 are calculated from sensor datathat were not corrected using the ozone cross-interference correction (Sect. 2.3.2) and thus represent an upper bound of the BL network uncertainty.
The Fig. 2 results and summary statistics are affected by a group of outliercollocations (unfilled circles in Fig. 2) that started in July 2019, duringwhich most sensors exhibited higher measurement error and poorer correlationto reference measurements. The eight collocations with the highest normalizedRMSE (>70 %) all occurred during July 2019 (Fig. S7).Additionally, seven of these July 2019 collocations hadR2 values below0.7, meaning they would have failed the statistical screening criteria usedfor determining valid collocation calibrations (Sect. 2.3.1). During thismonth, we observed high-biased sensor measurements when local airtemperatures were above 20–25 ∘C, which we discuss further below.With the July 2019 collocations (n=11) excluded, the median nRMSE and MBEof the remaining collocations (n=24) improve to 30 % and 4.5 µg m−3, respectively, and the medianR2 increases slightly to 0.81.
Figure 3 presents the collocation time series and monthly error statisticsbetween calibrated BL sensor and reference monitor measurements during twolong-term (>18 month) collocations, where the sensormeasurements are calibrated based on the collocation results during the2-week period directly preceding the extended evaluation period. While theaggregate MBE of both collocations is small (<2 µg m−3), BL sensors exhibit biases that vary seasonally relative toreference measurements; MBE of sensors during 14 d periods (Fig. 3b)ranges from−3 to+11 µg m−3 for unit 17(−8 % to+34 % of the 14 d mean concentration) and from−8 to+19 µg m−3 for unit 83 (−20 % to+91 % of the 14 d meanconcentration). The drifting sensor response follows the same seasonalpattern for both long-term collocations, with the highest bias occurringduring summer months and peaking during August 2020. Variations in RMSEerror are largely driven by sensor bias; nRMSE is highest during summermonths, corresponding to peak BL sensor bias. Figure S8 further illustratesthe occurrence of high-biased BL sensor measurements during hours when thelocal air temperature exceeded 20–25 ∘C. While the resultspresented above quantify uncertainty of sensors calibrated using referencecollocations, the data use cases in the following sections also includesensor data calibrated using two additional approaches when sensors couldnot be collocated at reference sites, as described in the “Methods” section (Sect. 2.3.1): transfer standard calibration and network calibration method. Thetransfer standard method is more difficult to validate because collocationsoccur at BL sites in the field instead of at reference sites. Theuncertainty of this method is expected to be marginally higher than thedirect reference site collocations in Fig. 2 due to the additional stepwhere the calibration is transferred between BL AQMesh units. A high levelof precision and consistency in response across BL NO2 sensors (R2=0.94, nRMSE = 0.1; Fig. S9) gives confidence that calibrations wouldtransfer effectively between units. Preliminary evaluations have shown thatthe estimated uncertainty of BL sensor measurements scaled with the networkcalibration method is broadly similar to the uncertainty of referencecollocation-calibrated sensors (Popoola et al., 2022; Popoola and Jones, 2020). The results in Figs. 2 and 3 demonstrate that the long-termmeasurement uncertainty of sensors calibrated during a brief, discreteperiod is influenced by the changes in sensor response during differentseasons and environmental conditions. Enhanced QA/QC such as calibration ona near-continuous basis or seasonal bias corrections such as the one shown in Fig. S10 (see Sect. 3.2.1) could minimize variations in measurement uncertaintydue to sensor performance.
For a long-term measurement campaign using sensors, evaluation againstreference measurements should be performed throughout the course of theproject. The evaluation results above point to the ability of BL sensors toaccurately reproduce changes in NO2 concentrations captured by thereference monitors (highR2 values) with average uncertainty (nRMSE) of∼ 35 %. However, our results also show that seasonal biasesdue to time-varying effects of environmental interferences can lead tolarger uncertainties (>100 % nRMSE) during periods when localair temperatures reached above 20–25 ∘C. This characterization ofsensor uncertainties can inform how results from the BL LC sensor networkare interpreted, ensuring derived insights are robust (e.g., differentiatingbetween high-biased sensor artifacts and elevated NO2 concentrations).
We next present a series of analytical use cases to evaluate theapplicability of BL NO2 LCS network results for deriving insights aboutthe local air pollution environment. Results from each use case using BLdata are compared against results generated from reference network data. Inaddition, the collocation sensor evaluations presented above are used toassess BL network uncertainty and interpret differences between BL andreference network results.
3.2 Use case validation
3.2.1 Regional pollution load and time trends
We first examine the ability of the BL sensor network to characterize trendsin the regional (Greater London) pollution load by comparing monthly meanNO2 concentrations of the BL network with the reference network results(Fig. 4). We compare monthly values here to assess the sensor network'sability to reproduce long-term patterns observed by the reference networkon timescales that would be sensitive to effects of seasonal variations inpollutant concentrations or sensor performance as well as long-term ambientpollution changes resulting from major interventions. We note two majorevents during the measurement campaign which are expected to impact NO2concentrations: (i) introduction of the Ultra Low Emission Zone (ULEZ), whichbecame effective on 8 April 2019, imposed tolls to discourage entry ofolder, higher-emitting vehicles into Central London, with increasingfractions of compliant vehicles and fewer vehicles overall observed in thezone through calendar year 2019 (GLA, 2020b), and (ii) Covid-19 pandemicrestrictions beginning March 2020, including social distancing measures andstay-at-home orders, disrupted activity patterns throughout Greater London.
Figure 4Comparison of monthly mean NO2 concentrations for the BL(n=100) and London reference (n=105) networks. Bottom panel showsdifference between networks. Vertical lines denote Ultra Low Emission Zone(ULEZ) start date (8 April 2019) and the start of the first Covid-19lockdowns (23 March 2020).
The BL network tracked the reference network trend while exhibiting lowermean concentrations for most of the campaign (on average 7 µg m−3 lower throughout campaign; 9 µg m−3 during2019). We attribute this partially to differences in location, sitetypes, and sampling points (height, distance to road, road traffic volume,etc.) between the networks, and this is confirmed through comparisons ofmeasured and modeled concentrations using the ADMS-Urban air pollutiondispersion model (described in Sect. 2.6). Modeled network mean NO2 concentrations for 2019 at reference network monitoring site receptorswere 5 µg m−3 higher (∼ 15 %) than the modeledmean concentrations at BL receptor locations. Because the model onlypredicts 55 % of the difference between the two networks, we examined themodel–network comparisons more closely. The model exhibits little systematicbias at reference sites (<1 µg m−3; see Fig. S11). Bycontrast, the mean of modeled concentrations was higher than that observedat the BL sites by 6 µg m−3, with the difference driven by BLsites with the lowest observed concentrations (Fig. S11). We note that 16 BLsites exhibited lower concentrations than the 20 µg m−3 minimumobserved by the reference network, so we cannot rule out the possibility oflow sensor bias in a portion of the BL network. In sum, we are unable tofully resolve the cause of the systematic difference between modeled andobserved BL concentrations, although it may have contributions fromuncertainty in sensor network measurements (and underlying QA/QC) and modeluncertainty.
Both networks show a downward year-on-year trend in NO2 concentrationsand seasonal variability with peak concentrations in the winter. However, BLNO2 means exhibit local maxima in July and August, when referencenetwork measurements are lowest. This effect is the most pronounced insummer 2020, which is the only time when the BL network average exceeds thatof the reference network. This bias of the BL network compared to referencenetwork trends during summer months is likely due in part to a systematichigh bias in the BL network's NO2 sensors coinciding with local airtemperatures above 20–25 ∘C, an effect which was evident duringcollocations with reference monitors (Figs. 3, S7, and S8). However,spatially varying NO2 pollution trends (e.g., Covid-19 restrictionshaving a larger impact on emissions at specific monitoring sites or cityneighborhoods) may have also affected the two networks differently andcontributed to the converging network means towards the end of the BLcampaign.
The long-term collocations (Sect. 3.1.2) were used to quantify seasonalchanges in sensor bias and could serve as a basis for an empiricalcorrection to the Fig. 4 BL network time series to improve the accuracy ofthe LCS results. This correction relies on the performance results beingconsistent across the network; the high precision between AQMesh units inour transfer standard collocations (medianR2=0.94; Fig. S9) supportsthis assumption for the BL project. In Fig. S10 we show the Fig. 4 BLtime series with a monthly bias correction based on the long-termcollocations that would largely mitigate the seasonal irregularities in theBL time series compared to the reference network.
3.2.2 Temporal pollution patterns
We next compare the recurring temporal patterns in NO2 concentrationsmeasured by the BL and reference networks (Fig. 5).
Figure 5Network mean diurnal and day-of-week NO2 concentrationpatterns in Greater London, as measured by the BL (n=99) and reference(n=105) networks during the pre-Covid-19 period of the BL project (1 October2018 through 29 February 2020). Bottom panel shows difference between networks.
The BL network captures diurnal and day-of-week patterns with three keydifferences from the reference network. First, BL network meanconcentrations are∼ 10 µg m−3 (23 %) lowerthan the reference network result. Most of this difference was predicted inthe modeling exercise discussed in Sect. 3.2.1, with additionalcontribution from uncertainty in sensor measurements. Second, BL networkmean concentrations show a reduced diurnal range compared to the regulatorynetwork (i.e., though daytime average BL concentrations are lower, nighttimevalues are similar to the reference network). This behavior may be due topreviously discussed differences in site characteristics (e.g., higher sensorplacement and lower traffic volume at near-road sites) yieldingreduced heterogeneity in site types across the BL network, which as a wholeappears to be measuring diurnal pollution patterns that are more in linewith urban background reference sites (Fig. 6). A similar effect is observedin a comparison of near-road (kerbside and roadside) and urban backgroundreference sites, where the concentration difference was smallest during late-night/early-morning hours (Fig. S12). A third key difference in diurnalday-of-week concentration patterns is the magnitude of the evening peak,which is consistently lower than the morning peak in the BL network. OnWednesdays, for example, the reference network evening peak reached 57 µg m−3 at 18:00 LT, while the BL network reached 40 µg m−3 at the same time; other weekdays similarly have the largestdifference in network mean concentrations during the evening rush hour peak.We have not identified a mechanism to explain this difference, which isevident, to a varying degree, throughout the year (Fig. S13).
Figure 6Weekday diurnal mean NO2 concentrations in Greater London asmeasured by the BL (n=70 near-road,n=40 urban background; number oflocations exceeds 100 because some devices were placed at multiple locationsduring the campaign) and reference (n=72 near-road,n=31 urbanbackground) networks during the pre-Covid-19 period of the BL project (1 October2018 through 29 February 2020) at two different site classification groups:near-road (left; includes sites classified as kerbside and roadside) andurban background (right). Bottom panel shows difference between networks.
The BL network was able to accurately characterize timing of peaks andtroughs in diurnal variability as well as capture differences in weekday andweekend pollution levels. Uncertainties in the precise magnitudes of somefeatures remain, with the evening peak registering relatively lower in theBL network.
3.2.3 Site type differences in diurnal pollution patterns
We next examine the ability of the BL network to detect differences indiurnal NO2 concentration patterns at different monitoring site types.Figure 6 shows the weekday diurnal averages for the BL and reference networkat near-road (kerbside and roadside) sites compared to urban backgroundsites.
In the morning, near-road concentrations peaked at 08:00–09:00 LT in both the BLand reference networks, reaching 60 µg m−3 in the referencenetwork and 50 µg m−3 in the BL network. The time of the eveningpeak was also consistent between networks, occurring at 18:00–19:00 LT andreaching 60 µg m−3 in the reference network compared to a lowerpeak of 44 µg m−3 in the BL network at the same time. In bothnetworks, the evening peak in concentrations occurred 1 h later (19:00–20:00 LT) at background sites than near-road sites. The greatest differencebetween BL and reference means at both near-road and urban background sitesoccurred during the evening peak in NO2 concentrations; this featureis identified in the network-wide trends in the prior section.
At this aggregate level, the lower-cost network captures similar diurnalfeatures and effectively differentiates between pollution levels andtime-of-day trends at urban background and near-road sites.
3.2.4 Hotspots and spatial heterogeneity
Here we discuss the application of BL LCS data for identifying hotspots andcharacterizing spatial heterogeneity in NO2 concentrations, using acase study where BL sensor measurements led to identification of an airpollution hotspot. During the first winter of the project (December 2018 throughFebruary 2019), a BL sensor deployed at Holloway Bus Garage measured mean weekdayNO2 concentrations of 77 µg m−3, 89 % higher than the BLnetwork weekday mean of 41 µg m−3 (Fig. S14). Though theconcentration gradient (between Holloway Bus Garage and the BL network mean)was larger than the typical sensor uncertainties (∼ 35 %nRMSE) and occurred during winter months when large positive biases werenot observed during collocation evaluations, additional steps were taken toestablish confidence that the local pollution levels were accuratelycharacterized and not sensor artifacts. Two additional BL sensors weredeployed in the area, and a follow-up transfer standard collocation wasperformed which verified the accuracy of the deployed pod's calibrationfactors.
The BL monitoring at Holloway Bus Garage ultimately led to corrective actionby local authorities, and this successful example demonstrates the potentialvalue of LCSs for identifying air pollution hotspots. The case study alsoemphasizes the need for rigorous verification of measurements from anindividual sensor. The collocation analyses quantified a wide range in thebias of BL sensors over the course of the project as well as uncertainty inthe consistency of sensor performance over time (Figs. 2 and 3). Therefore,especially for concentration gradients of similar magnitude to the estimateduncertainty of the sensors, there is a need for caution when analyzingsite-specific data; we established confidence in the LC sensor hotspotcharacterization through the deployment of additional LC sensors to verifyresults.
3.2.5 High-pollution episodes
Here we test the viability of the BL network to detect short- to medium-term(hours to days) episodes of elevated NO2 concentrations using awell-characterized historical air pollution event in December 2019 (LAQN,2019). Weather conditions in Greater London resulted in the formation of astrong temperature inversion that caused a build-up of primary pollutants,including NO2, in the layer of colder air close to the ground, withpollution peaking at morning rush hour on 4 December (LAQN, 2019).Figure 7 compares the hourly mean NO2 concentrations as measured by theBL and reference networks for the week of the pollution episode.
Figure 7Hourly network mean NO2 concentrations for the BL(n=85) and reference (n=105) networks during a high-pollution episode inDecember 2019. Bottom panel shows difference between networks.
The BL network detected a short-term regional build-up of pollution with atemporal profile that provides excellent comparability with the referencenetwork result (R2=0.96) and corresponds to the London Air QualityNetwork's published report about the event. The highest peak occurred duringlate rush hour (09:00–10:00 LT) on the morning of 4 December, with the BLnetwork registering a peak of 87 µg m−3 compared to 103 µg m−3 for the reference network during the same time period. Anothersmaller peak occurred when evening emissions were trapped on 4 December,and the event subsided when the inversion broke near midnight on 4 December. The BL lower-cost network captures the basic features of the event,although there is a low bias compared to the reference sites (nRMSE=23 %, MBE=−11 µg m−3), partially explained by thedifferent site types (see Sect. 3.2.1).
However, we note that the network was less effective in characterizingpollution events during periods of poorer sensor performance. In Fig. S15 wepresent a more cautionary case study during July 2019. We demonstrated inSect. 3.1.2 that collocated BL sensors produced high-biased measurementsduring periods when local air temperatures reached above 20–25 ∘C,with worst-case nRMSE exceeding 100 % (dominated by positive bias). Figure S15 shows an instance where this effect leads to an overestimation ofregional NO2 pollution levels using the BL network (for example, the BL networkmean during daytime hours on 25 July 2019 is 91 µg m−3 compared to the reference network mean of 65 µg m−3, a 40 %positive bias across the network). Due to the extensive reference network inLondon and frequent short-term as well as ongoing long-term BL sensorcollocations, we were able to identify the apparently anomalous BL sensorbehavior under these environmental conditions which resulted in a systematicpositive bias across the network. However, in cases with limited referencemonitoring infrastructure, the BL measurements could have led to anoverestimation of the magnitude of the pollution event in question. Whiletechnological and methodological solutions to address this sensor issue areviable, another project with different technologies or environmentalcharacteristics may experience different effects, illustrating theimportance of rigorous data validation and uncertainty evaluation in thecontext of each new application of LCS technology.
In a LCS deployment, careful evaluation of sensor performance (which mayvary between projects due to, for example, specific sensor technology, firmware,local meteorology and pollution characteristics, among others) maximizes thevalue of the data by informing how they should be processed, analyzed, andinterpreted. Robust uncertainty characterization and validation againstreference instruments equips the user to take full advantage of data,including (i) developing corrections (see Fig. S10 presenting the BL networktime series with a potential correction derived from collocation evaluationresults), (ii) excluding measurements during conditions where sensorperformance might be compromised, or (iii) ensuring analyses are appropriatebased on the data quality. By contrast, we have shown that without adetailed understanding of variations in sensor performance across a campaign(see Fig. 3b illustrating temperature-related drift), biased sensormeasurements at some moments during the project could have led users tooverestimate pollution levels or, over longer timeframes, miss trends inconcentration patterns. Our findings emphasize the importance of monitoringsensor performance for the duration of a measurement campaign as even pre-and post-campaign sensor evaluations may not have detected the seasonalchanges in sensor performance that our repeat (Fig. 2) and long-term (Fig. 3) collocations allowed us to quantify. A near-real-time calibrationapproach may also be valuable for tracking and improving sensor performanceover time by providing continuous calibrations and assessments of networkperformance, although a single point calibration was used here (Considine et al.,2021). Our results also demonstrate how LC sensors could be used in a citywith more limited existing monitoring infrastructure than in London. The BLnetwork generated a series of insights about air pollution in London that wecompared to reference network results, and we found that the BL LCS networkcharacterized many NO2 trends and patterns effectively, includingyear-over-year concentration trends, timing of diurnal peaks,weekday–weekend concentration gradients, and profiles of short- tomedium-term periods of elevated pollution. We also showed how BL sensoruncertainties, which were evaluated using collocations at three Londonreference monitors, limited the LCS network's ability to capture preciselysome features of air pollution trends, emphasizing that especially in aplace without an extensive reference network, it is advisable to have atleast one reliable reference instrument as a basis for ongoing LC sensorcalibration and uncertainty evaluation. We also note that the use ofrepresentative reference collocations (i.e., keeping one or two units atreference sites throughout the project) to estimate network performancerelies on the testable assumption that sensors are highly precise across thenetwork.
The sensor uncertainties and data use cases that we have evaluated arespecific to the sensor technology and firmware used as well as the localenvironmental characteristics in London. In London, environmental effectssignificantly impacted data quality, including frequent wintertime sensorfailures and high measurement artifacts occurring when local ambient airtemperatures exceeded 20–25 ∘C, indicating that sensor performancecould vary in other cities with different source patterns and meteorology.Furthermore, in another environment with different air pollution levels, thesame magnitude of sensor RMSE may represent a different proportion of theaverage concentration, reinforcing the need to evaluate sensor performancelocally and consider the tolerable amount of measurement error for eachapplication. Additionally, the current absence of a performance standard forLC sensors exposes the end user to risks in the sensor selection process,making it advisable for each implementation of LCS technology to perform itsown performance evaluation. Our approach can provide a roadmap for futureLCS deployments to maximize data quality and confidence in resultinginsights by following robust QA/QC protocols, most notably the tracking ofrepresentative sensor performance for the duration of the project via directtraceability to reference measurements.
BL network data are available on the OpenAQ platform:https://openaq.org/#/project/28967 (Breathe London, 2022). London reference monitor data can be accessed usingthe R openair package (Carslaw and Ropkins, 2012). The corresponding data can also be accessed throughhttps://www.londonair.org.uk/london/asp/datadownload.asp (LondonAir, 2022). Meteorological data are available from the NOAA Integrated Surface Database (https://www.ncdc.noaa.gov/isd; NOAA, 2021) and can be accessed via the R worldmet package (https://CRAN.R-project.org/package=worldmet; Carslaw, 2020).Collocation data are available upon request from the corresponding author.
The supplement related to this article is available online at: https://doi.org/10.5194/amt-15-321-2022-supplement.
ERF, MDT, FD, and KM managed the project, including coordinating sensordeployment and collocations. JM oversaw installation, operation, andmaintenance of the sensor network. NAM hosted sensor system deployment. RLJand OAMP developed and applied the remote network calibration method and ozonecorrection to the hourly NO2 dataset and were involved in datacuration. AS, ERF, and DC applied the QA/QC procedure to the raw1 min AQMesh measurements to produce a calibrated hourly dataset, curateddatasets and metadata, and developed and ran model simulations. RAA, RLJ, DC,and NAM supervised research. RAA, DRP, and LEP formulated research goals forthis paper. DRP prepared the manuscript, including formal analysis,visualization, and writing. All co-authors contributed to reviewing andediting.
During parts of the Breathe London pilot project, Katie Moore and Jim Mills were employed atcommercial sensor providers (Clarity Movement Co. and ACOEM Air MonitorsLtd., respectively), and the University of Cambridge had a commercialarrangement with AQMesh; these relationships did not affect the workpresented here. All other authors declare they have no competing interests.
Publisher’s note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
The Breathe London pilot project was convened by C40 Citiesand the Mayor of London. The authors would like to thank the many hosts ofBreathe London monitors, including local councils, schools, and residents, aswell as the scientific and project advisors for their contributions. Theauthors are especially grateful to the local councils of Camden, Southwark,and Islington for continued access to reference monitors for collocationsthat were critical to this study. Thanks also to Greg Slater for his inputon data visualization.
This work was supported by the Children's Investment Fund Foundation, withcontinued funding from the Clean Air Fund (grant numbers 1908-03995 and 341) and further fundingsupport provided by Signe Ostby and Scott Cook (of the Valhalla CharitableFoundation) as well as funding by the Mayor of London for 10 additionalAQMesh units.
This paper was edited by Dominik Brunner and reviewed by Laurent Spinelle and one anonymous referee.
Apte, J. S., Messier, K. P., Gani, S., Brauer, M., Kirchstetter, T. W.,Lunden, M. M., Marshall, J. D., Portier, C. J., Vermeulen, R. C. H., andHamburg, S. P.: High-resolution Air Pollution Mapping with Google StreetView Cars: Exploiting Big Data, Environ. Sci. Technol., 51, 6999–7008,https://doi.org/10.1021/acs.est.7b00891, 2017.
AQMesh:https://www.aqmesh.com/products/aqmesh/, last access:15 June 2021.
AQ-SPEC: AQMesh (v.4.0) – field evaluation, South Coast AQMD, available at:http://www.aqmd.gov/aq-spec/sensordetail/aqmesh-(v.4.0) (last access: 7 January 2022), Diamond Bar, CA,2015.
Bi, J., Stowell, J., Seto, E. Y. W., English, P. B., Al-Hamdan, M. Z.,Kinney, P. L., Freedman, F. R., and Liu, Y.: Contribution of low-cost sensormeasurements to the prediction of PM2.5 levels: A case study in ImperialCounty, California, USA, Environ. Res., 180, 108810,https://doi.org/10.1016/j.envres.2019.108810, 2020.
Breathe London: AQMesh fixed sensor network data quality assurance andcontrol procedures, available at:https://www.globalcleanair.org/files/2021/01/Breathe-London-Fixed-Sensor-Network-QAQC-Procedures.pdf (last access: 7 January 2022),2020.
Breathe London: Breathe London archival website, available at:http://breathelondon.edf.org/, last access: 15 June 2021, 2021a.
Breathe London: The Breathe London Blueprint, available at:https://www.globalcleanair.org/files/2021/02/EDF-Europe-BreatheLondon_Blueprint-guide.pdf (last access: 7 January 2022), 2021b.
Breathe London: Breathe London Stationary, OpenAQ [data set], available at:https://openaq.org/#/project/28967, last access: 7 January 2022.
Carruthers, D., Stidworthy, A., Clarke, D., Dicks, J., Jones, R., Leslie,I., Popoola, O. A. M., and Seaton, M.: Urban emission inventory optimisationusing sensor data, an urban air quality model and inversion techniques,Int. J. Environ. Pollut., 66, 252,https://doi.org/10.1504/IJEP.2019.104878, 2019.
Carslaw, D.: worldmet: Import Surface Meteorological Data from NOAAIntegrated Surface Database (ISD), R package version 0.9.2, available at:https://CRAN.R-project.org/package=worldmet (last access: 7 January 2022), 2020.
Carslaw, D. and Ropkins, K.: openair – An R package for air quality dataanalysis, Environ. Model. Softw., 27–28, 52–61,https://doi.org/10.1016/j.envsoft.2011.09.008, 2012.
Castell, N., Dauge, F. R., Schneider, P., Vogt, M., Lerner, U., Fishbain,B., Broday, D., and Bartonova, A.: Can commercial low-cost sensor platformscontribute to air quality monitoring and exposure estimates?, Environ.Int., 99, 293–302,https://doi.org/10.1016/j.envint.2016.12.007,2017.
Castell, N., Schneider, P., Grossberndt, S., Fredriksen, Mirjam. F.,Sousa-Santos, G., Vogt, M., and Bartonova, A.: Localized real-timeinformation on outdoor air quality at kindergartens in Oslo, Norway usinglow-cost sensor nodes, Environ. Res., 165, 410–419,https://doi.org/10.1016/j.envres.2017.10.019, 2018.
Caubel, J. J., Cados, T. E., Preble, C. V., and Kirchstetter, T. W.: ADistributed Network of 100 Black Carbon Sensors for 100 Days of Air QualityMonitoring in West Oakland, California, Environ. Sci. Technol., 53,7564–7573,https://doi.org/10.1021/acs.est.9b00282, 2019.
CERC: Final report, Breathe London project, available at:https://www.globalcleanair.org/files/2021/02/BL-CERC-Final-Report.pdf (last access: 7 January 2022), 2021.
Clark, L. P., Millet, D. B., and Marshall, J. D.: National Patterns inEnvironmental Injustice and Inequality: Outdoor NO2 Air Pollution in theUnited States, PLOS ONE, 9, e94431,https://doi.org/10.1371/journal.pone.0094431, 2014.
Considine, E. M., Reid, C. E., Ogletree, M. R., and Dye, T.: Improving accuracy of air pollution exposure measurements: Statistical correction of a municipal low-cost airborne particulate matter sensor network, Environ. Pollut., 268, 115833,https://doi.org/10.1016/j.envpol.2020.115833, 2021.
Dajnak, D., Evangelopoulos, D., Kitwiroon, N., Beevers, S. D., and Walton,H.: London health burden of current air pollution and future health benefitsof mayoral air quality policies, available at:http://erg.ic.ac.uk/research/home/resources/ERG_ImperialCollegeLondon_HIA_AQ_LDN_11012021.pdf (last access: 7 January 2022), 2021.
Duvall, R. M., Long, R. W., Beaver, M. R., Kronmiller, K. G., Wheeler, M.L., and Szykman, J. J.: Performance Evaluation and Community Application ofLow-Cost Sensors for Ozone and Nitrogen Dioxide, Sensors, 16, 1698,https://doi.org/10.3390/s16101698, 2016.
EU: Directive 2008/50/EC of the European Parliament and of the Council of 21May 2008 on ambient air quality and cleaner air for Europe, Off. J. Eur.Union, 152, 1–44, 2008.
Greater London Authority (GLA): Guide for monitoring air quality in London,available at:https://www.london.gov.uk/sites/default/files/air_quality_monitoring_guidance_january_2018.pdf (last access: 7 January 2022), 2018.
Greater London Authority (GLA): Air pollution monitoring data in London:2016 to 2020, available at:https://www.london.gov.uk/sites/default/files/air_pollution_monitoring_data_in_london_2016_to_2020_feb2020.pdf (last access: 7 January 2022), 2020a.
Greater London Authority (GLA): Central London ultra low emission zone –ten month report, available at:https://www.london.gov.uk/sites/default/files/ulez_ten_month_evaluation_report_23_april_2020.pdf (last access: 7 January 2022),2020b.
Greater London Authority (GLA): Monitoring and predicting air pollution,available at:https://www.london.gov.uk/what-we-do/environment/pollution-and-air-quality/monitoring-and-predicting-air-pollution (last access: 7 January 2022),2021.
Gupta, P., Doraiswamy, P., Levy, R., Pikelnaya, O., Maibach, J., Feenstra,B., Polidori, A., Kiros, F., and Mills, K. C.: Impact of California Fires onLocal and Regional Air Quality: The Role of a Low-Cost Sensor Network andSatellite Observations, GeoHealth, 2, 172–181,https://doi.org/10.1029/2018GH000136,2018.
Health Effects Institute (HEI): State of Global Air 2020, Health EffectsInstitute, Boston, MA, 2020.
Heimann, I., Bright, V. B., McLeod, M. W., Mead, M. I., Popoola, O. A. M.,Stewart, G. B., and Jones, R. L.: Source attribution of air pollution byspatial scale separation using high spatial density networks of low cost airquality sensors, Atmos. Environ., 113, 10–19,https://doi.org/10.1016/j.atmosenv.2015.04.057, 2015.
Jiao, W., Hagler, G., Williams, R., Sharpe, R., Brown, R., Garver, D., Judge, R., Caudill, M., Rickard, J., Davis, M., Weinstock, L., Zimmer-Dauphinee, S., and Buckley, K.: Community Air Sensor Network (CAIRSENSE) project: evaluation of low-cost sensor performance in a suburban environment in the southeastern United States, Atmos. Meas. Tech., 9, 5281–5292,https://doi.org/10.5194/amt-9-5281-2016, 2016.
Karagulian, F., Barbiere, M., Kotsev, A., Spinelle, L., Gerboles, M.,Lagler, F., Redon, N., Crunaire, S., and Borowiak, A.: Review of thePerformance of Low-Cost Sensors for Air Quality Monitoring, Atmosphere, 10, 506,https://doi.org/10.3390/atmos10090506, 2019.
Kelly, K. E., Whitaker, J., Petty, A., Widmer, C., Dybwad, A., Sleeth, D.,Martin, R., and Butterfield, A.: Ambient and laboratory evaluation of alow-cost particulate matter sensor, Environ. Pollut., 221, 491–500,https://doi.org/10.1016/j.envpol.2016.12.039, 2017.
Kim, J., Shusterman, A. A., Lieschke, K. J., Newman, C., and Cohen, R. C.: The BErkeley Atmospheric CO2 Observation Network: field calibration and evaluation of low-cost air quality sensors, Atmos. Meas. Tech., 11, 1937–1946,https://doi.org/10.5194/amt-11-1937-2018, 2018.
LondonAir: Data Downloads, Imperial College London [data set], available at:https://www.londonair.org.uk/london/asp/datadownload.asp, last access: 7 January 2022.
London Air Quality Network (LAQN): LAQN Pollution Episodes, available at:https://londonair.org.uk/london/asp/publicepisodes.asp?region=0&site=&postcode=&la_id=&level=All&bulletindate=03%2F12%2F2019&MapType=Google&zoom=&lat=51.4750&lon=-0.119824&VenueCode=&bulletin=explanation&episodeID=pol3to4Dec2019&pageID=page1&cm-djitdk-djitdk=(last access 20 January 2021), 2019.
Lewis, A. C., Lee, J. D., Edwards, P. M., Shaw, M. D., Evans, M. J., Moller,S. J., Smith, K. R., Buckley, J. W., Ellis, M., Gillot, S. R., and White,A.: Evaluating the performance of low cost chemical sensors for airpollution research, Faraday Discuss., 189, 85–103,https://doi.org/10.1039/C5FD00201J, 2016.
Lopez-Restrepo, S., Yarce, A., Pinel, N., Quintero, O. L., Segers, A., andHeemink, A. W.: Urban Air Quality Modeling Using Low-Cost Sensor Network andData Assimilation in the Aburrá Valley, Colombia, Atmosphere, 12, 91,https://doi.org/10.3390/atmos12010091, 2021.
McHugh, C. A., Carruthers, D. J., and Edmunds, H. A.: ADMS–Urban: an airquality management system for traffic, domestic and industrial pollution, Int. J. Environ. Pollut., 8,666–674, 1997.
Mead, M. I., Popoola, O. A. M., Stewart, G. B., Landshoff, P., Calleja, M.,Hayes, M., Baldovi, J. J., McLeod, M. W., Hodgson, T. F., Dicks, J., Lewis,A., Cohen, J., Baron, R., Saffell, J. R., and Jones, R. L.: The use ofelectrochemical sensors for monitoring urban air quality in low-cost,high-density networks, Atmos. Environ., 70, 186–203,https://doi.org/10.1016/j.atmosenv.2012.11.060, 2013.
Miller, D. J., Actkinson, B., Padilla, L., Griffin, R. J., Moore, K., Lewis,P. G. T., Gardner-Frolick, R., Craft, E., Portier, C. J., Hamburg, S. P.,and Alvarez, R. A.: Characterizing Elevated Urban Air Pollutant SpatialPatterns with Mobile Monitoring in Houston, Texas, Environ. Sci. Technol.,54, 2133–2142,https://doi.org/10.1021/acs.est.9b05523, 2020.
Munir, S., Mayfield, M., Coca, D., Jubb, S. A., and Osammor, O.: Analysingthe performance of low-cost air quality sensors, their drivers, relativebenefits and calibration in cities-a case study in Sheffield, Environ. Monit.Assess., 191, 94,https://doi.org/10.1007/s10661-019-7231-8, 2019.
NOAA: Integrated surface database (ISD), available at:https://www.ncdc.noaa.gov/isd (last access: 7 January 2022), 2021.
Pinder, R. W., Klopp, J. M., Kleiman, G., Hagler, G. S. W., Awe, Y., andTerry, S.: Opportunities and challenges for filling the air quality data gapin low- and middle-income countries, Atmos. Environ., 215, 116794,https://doi.org/10.1016/j.atmosenv.2019.06.032, 2019.
Pope, F. D., Gatari, M., Ng'ang'a, D., Poynter, A., and Blake, R.: Airborne particulate matter monitoring in Kenya using calibrated low-cost sensors, Atmos. Chem. Phys., 18, 15403–15418,https://doi.org/10.5194/acp-18-15403-2018, 2018.
Popoola, O. A. M., Carruthers, D., Lad, C., Bright, V. B., Mead, M. I.,Stettler, M. E. J., Saffell, J. R., and Jones, R. L.: Use of networks of lowcost air quality sensors to quantify air quality in urban settings,Atmos. Environ., 194, 58–70,https://doi.org/10.1016/j.atmosenv.2018.09.030, 2018.
Popoola, O. A. M. and Jones, R. L.: A novel calibration method forhyperlocal measurements of air quality using a low-cost sensor network, AirSensors International Conference (ASIC): Virtual Fall Series, October 2020,available at:https://www.youtube.com/watch?v=sPzwmLNiP1w&ab_channel=UCDavisAirQualityResearchCenter (last access: 7 January 2022), 2020.
Popoola, O. A. M., Fleming, J., Peters, D. R., Alvarez, R. A., Ma, G., Stidworthy, A., Forsyth, E., Martin, N. A., Mills, J., Carruthers, L. E., Fonseca, E. R., and Jones, R. L.: A cloud based calibration method for atmospheric measurement networks, in preparation, 2022.
Sahu, R., Nagal, A., Dixit, K. K., Unnibhavi, H., Mantravadi, S., Nair, S., Simmhan, Y., Mishra, B., Zele, R., Sutaria, R., Motghare, V. M., Kar, P., and Tripathi, S. N.: Robust statistical calibration and characterization of portable low-cost air quality monitoring sensors to quantify real-time O3 and NO2 concentrations in diverse environments, Atmos. Meas. Tech., 14, 37–52,https://doi.org/10.5194/amt-14-37-2021, 2021.
Shah, R. U., Robinson, E. S., Gu, P., Apte, J. S., Marshall, J. D.,Robinson, A. L., and Presto, A. A.: Socio-economic disparities in exposureto urban restaurant emissions are larger than for traffic, Environ. Res.Lett., 15, 114039,https://doi.org/10.1088/1748-9326/abbc92, 2020.
Spinelle, L., Gerboles, M., Villani, M. G., Aleixandre, M., andBonavitacola, F.: Field calibration of a cluster of low-cost availablesensors for air quality monitoring. Part A: Ozone and nitrogen dioxide,Sensors and Actuators B: Chemical, 215, 249–257,https://doi.org/10.1016/j.snb.2015.03.031, 2015.
The Guardian: UK has broken air pollution limits for a decade, EU courtfinds, available at:https://www.theguardian.com/environment/2021/mar/04/uk-has-broken-air-pollution-limits-for-a-decade-eu-court-finds (last access: 7 January 2022),2021.
US GAO: Air pollution: Opportunities to better sustain and modernize thenational air quality monitoring system, Washington, D.C., GAO-21-38, 2020.
WHO: Ambient (outdoor) air pollution, available at:https://www.who.int/news-room/fact-sheets/detail/ambient-(outdoor)-air-quality-and-health(last access: 28 December 2020), 2018.
WMO: An Update on Low-cost Sensors for the Measurement of AtmosphericComposition, Geneva, Switzerland, WMO-No. 1215, 2021.
Zimmerman, N., Presto, A. A., Kumar, S. P. N., Gu, J., Hauryliuk, A., Robinson, E. S., Robinson, A. L., and R. Subramanian: A machine learning calibration model using random forests to improve sensor performance for lower-cost air quality monitoring, Atmos. Meas. Tech., 11, 291–313,https://doi.org/10.5194/amt-11-291-2018, 2018.