Low-cost Particle Sensors Research Articles

Aerosol sources characterization and apportionment from low-cost particle sensors in an urban environment

Low-cost sensors (LCS) have the potential to provide accurate and reliable measurements of air quality in real-time. This improves our ability to monitor, identify sources of pollution and develop mitigation strategies for effective air quality management. However, recent research on LCS has primarily focused on monitoring, exposure assessment, and calibration. In this study, we investigate the applicability of LCS data collected at ambient sites for characterizing and apportioning aerosol sources. Non-negative matrix factorization (NMF) was applied to the size-resolved data collected across five sites within the Indian Institute of Technology Bombay (IITB) campus in Mumbai using the LCS Alphasense OPC-N2. The sampling was done for 15 days at 5 locations in IITB, and each site only had 3 days of data. NMF resolved two factors for three sites, namely aromas (S2), hostel hub (S3) and central library (S4), while three factors were resolved for two sites, namely construction site (S1) and main gate (S5). Two common sources were determined for all the sites: (i) dust and marine source and (ii) traffic and combustion sources, which agree with the sources identified by studies in the literature. The third factor resolved at sites S1 and S5 is representative of heavy-duty diesel vehicles (HDDVs), which is present for a very short period and is captured because of the capability of high temporal resolution of the LCS. This offers a unique, cost-effective advantage of LCS for capturing episodic activities. The study suggests that in low- and middle-income countries with limited air quality monitoring capabilities, the size-time-resolved PM concentration data obtained from a network of low-cost sensors can estimate the pollution sources. This study provided evidence that despite their inherent limitations, LCS can be useful in attaining interpretable information about pollution sources and recommends extensive use of LCS for source characterization in the future.

Health risks for children exercising in an air-polluted environment can be reduced by monitoring air quality with low-cost particle sensors

A child’s body is highly sensitive to air quality, especially regarding the concentration of particulate matter (PM). Nevertheless, due to the high cost of precision instruments, measurements of PM concentrations are rarely carried out in school areas where children spend most of their daily time. This paper presents the results of PM measurements made by a validated, low-cost university air pollution measurement system operating in a rural area near schools. An assessment of children’s exposure to PM during school hours (8 a.m.–6 p.m.) at different times of the year was carried out. We show that PM10 concentrations in the air, particularly in winter, often exceeded the alert values of 50 µg m−3, posing a health risk to children, especially when children exercise outside the school building. We also calculated the rate and total PM10 deposition in the respiratory tract during various physical activities performed in clean and polluted air. Monitoring actual PM10 concentrations as presented in this paper, using a low cost sensors, offer school authorities and teachers an opportunity to reduce health risks for children. This can be achieved by adjusting the duration and exercise intensity of children’s outdoor physical activities according to the measured air quality.

Validity of a Low-cost Particle Sensor to Assess Real-time Indoor PM2.5 Exposure

Validity of a Low-cost Particle Sensor to Assess Real-time Indoor PM2.5 Exposure

Long-term application of low-cost sensors for monitoring indoor air quality and particle dynamics in a commercial building

Human exposure to indoor particles has been associated with increased rates of morbidity and mortality. To develop efficient particle control measures in indoor environments, there has been an increasing demand for the use of low-cost particle sensors towards indoor air quality applications. However, few studies have examined the potential application of low-cost particle sensors for monitoring indoor air quality and particle dynamics under a long-term deployment in commercial buildings. Given this background, we conducted a case study of six-month field investigation on the use of low-cost sensors for evaluating particle emission, concentration, and removal in an office building. Our results illustrate the necessity of improving control strategies for fine particles <1 μm in commercial buildings (e.g., utilizing high efficiency particle filters). Deploying low-cost particle sensors within an air handling unit and occupied spaces could be a quick and simple approach for monitoring particle filtration effectiveness of a mechanical ventilation system. Furthermore, this study suggests that coupling low-cost particle sensors with CO2 measurements can offer informative data for analyzing particle emissions associated with occupant activities. Our results reveal that particle resuspension due to occupants can be an important emission source for particles >5 μm in office environments. The study findings could help future research and applications of low-cost sensors in monitoring indoor air quality in commercial buildings.

Long-Term Analysis of Aerosol Concentrations Using a Low-Cost Sensor: Monitoring African Dust Outbreaks in a Suburban Environment in the Canary Islands.

This study presents the results of the long-term monitoring of PM10 and PM2.5 concentrations using a low-cost particle sensor installed in a suburban environment in the Canary Islands. A laser-scattering Nova Fitness SDS011 sensor was operated continuously for approximately three and a half years, which is longer than most other studies using this type of sensor. The impact of African dust outbreaks on the aerosol concentrations was assessed, showing a significant increase in both PM10 and PM2.5 concentrations during the outbreaks. Additionally, a good correlation was found with a nearby reference instrument of the air quality network of the Canary Islands' government. The correlation between the PM10 and PM2.5 concentrations, the effect of relative humidity, and the stability of the sensor were also investigated. This study highlights the potential of this kind of sensor for long-term air quality monitoring with a view to developing extensive and dense low-cost air quality networks that are complementary to official air quality networks.

Cracking the code—Matching a proprietary algorithm for a low-cost sensor measuring PM1 and PM2.5

Some manufacturers of low-cost particle sensors use proprietary algorithms to estimate particle mass concentrations such as PM2.5. Often little or no information is given regarding the calibration aerosol, how the algorithm was created or tested, or how the mass was estimated from the particle number counts. If the algorithm is faulty in some way, researchers have little ability to correct it in a fundamental way, although they can multiply the output by some calibration factor to match the particular aerosol combination they are studying. However, the adjustment still requires the use of the proprietary algorithm, which may have quirks that make it impossible to fix completely using a single calibration factor. It might be possible in some cases to avoid using the proprietary algorithm at all. That is the approach of this study. The low-cost sensor studied is the Plantower PMS 5003, and the algorithm is the CF_1 algorithm offered by the manufacturer. Data from a six-month study of four collocated PurpleAir PA-II monitors, each containing two independent Plantower PMS 5003 sensors, were collected. Two of these monitors had previously been calibrated against research-grade monitors. The best-fitting model for PM1 was found to be of the form PM1 = a*(N1 + N2) + d, where N1 and N2 are the particle numbers in the size categories 0.3–0.5 μm and 0.5–1 μm, and d is an additive constant. The best-fitting model for PM2.5 was of the form a*(N1 + N2) + b*N3 + d, where N3 is the number of particles in the third size fraction (1–2.5 μm). The individual models for all 8 sensors matched the reported CF_1 values for both PM1 and PM2.5 with R2 values exceeding 0.99, intercepts near zero, and slopes in the 0.99–1.01 range. The proposed models may also explain why the CF_1 algorithm reports values of zero for a substantial portion of PM1 and PM2.5 estimates. General models capable of being applied to other datasets were developed and estimated to have mean absolute errors (MAEs) <1 μg/m3.

Aerosol Measurement Degradation in Low-Cost Particle Sensors Using Laboratory Calibration and Field Validation.

Increasing concern over air pollution has led to the development of low-cost sensors suitable for wide-scale deployment and use by citizen scientists. This project investigated the AirU low-cost particle sensor using two methods: (1) a comparison of pre- and post-deployment calibration equations for 24 devices following use in a field study, and (2) an in-home comparison between 3 AirUs and a reference instrument, the GRIMM 1.109. While differences (and therefore some sensor degradation) were found in the pre- and post-calibration equation comparison, absolute value changes were small and unlikely to affect the quality of results. Comparison tests found that while the AirU did tend to underestimate minimum and overestimate maximum concentrations of particulate matter, ~88% of results fell within ±1 μg/m3 of the GRIMM. While these tests confirm that low-cost sensors such as the AirU do experience some sensor degradation over multiple months of use, they remain a valuable tool for exposure assessment studies. Further work is needed to examine AirU performance in different environments for a comprehensive survey of capability, as well as to determine the source of sensor degradation.

Significance of sources and size distribution on calibration of low-cost particle sensors: Evidence from a field sampling campaign

Low-cost sensors (LCS) are gathering the interest of researchers and monitoring agencies worldwide due to their compact size and economic feasibility. However, the data recorded by LCS is often of low quality owing to its calibration dependencies and biases. An intensive field sampling campaign was conducted at five sites inside the IIT Bombay campus to understand the fundamental issues associated with LCS. PA-II and OPC-N2 LCS were collocated with BAM-1020 and OPS- 3330 at the sampling locations for inter-comparison. LCS shows a good correlation with BAM-1020 at sites where the contributing sources produce more particles in the 1–2.5 μm size range than particles in below 1 μm (PM1) and 2.5–10 μm size range. The performance decreased with an increase in mass fractions of PM1 and PM10. The overall performance of both PA-II (R2 = 0.72) and OPC-N2 (R2 = 0.73) are comparable. Both PA-II and OPC-N2 have substandard performance with R2 in the range of 0.30–0.39 and 0.42–0.53 at the construction and main gate site respectively. Comparing the two calibration approaches used in this study indicates the importance of including size distribution parameters in the calibration of LCS. The calibration models were developed for each site and were compared with the general model developed for PA-II and OPC-N2. Results indicate that the site-specific models are in better agreement with the reference instrument than the general calibration model. The number concentration recorded by PA-II was poorly correlated with OPS-3330, especially for particles >1 μm and vice versa for OPC-N2. The particle count for PM > 2.5 μm recorded by PA-II is predominantly zero, which is inconsistent with the mass concentration data recorded by the sensor. The size distribution results indicate that LCS assumes a universal monotonically decreasing function of number concentration with respect to the particle diameter. It is one of the critical problems with LCS measurements as any error in the number measurement is increased 3-fold in the mass conversion. This study shows the need for site-specific robust calibration of LCS based on the particle size distribution and provides a direction in their development.

Barometric Pressure as a correction factor for low-cost particulate matter sensors

Concentration of particulate matter constitutes a major, environmental problem affecting human health especially in urban areas. Although research centers and government agencies have air quality monitoring systems installed, their limited number prevents the spatiotemporal analysis in urban areas. EU funds research projects to engage citizens for air quality issues. The aim is to involve citizens to install low-cost air quality measuring stations to develop a wider network and to retrieve accurate information about air quality. The use of low-cost sensors arise questions about the reliability of the measurements. It is known, that particles concentration measurement are affected by humidity and temperature. Several models have been proposed to remove the impact of the temperature and humidity on the particle’s measurements but the impact of barometric pressure is under investigation. In this work, an extensive study is performed on the barometric pressure aftermath and corrective factors are proposed in relation to it. The pilot stations were installed next to high cost official stations for a month to calibrate the low-cost sensors. Herein, a model of multiple linear regression is presented to confirm the importance of barometric pressure correction, with the ultimate goal being the optimization of the low-cost particle sensors measured values.

Long-Term Trends in Secondhand Smoke Exposure in High-Rise Housing Serving Low-Income Residents in New York City: Three-Year Evaluation of a Federal Smoking Ban in Public Housing, 2018-2021.

In July 2018, the U.S. Department of Housing and Urban Development passed a rule requiring public housing authorities to implement smoke-free housing (SFH) policies. We measured secondhand smoke (SHS) exposure immediately before, and repeatedly up to 36 months post-SFH policy implementation in a purposeful sample of 21 New York City (NYC) high-rise buildings (>15 floors): 10 NYC Housing Authority (NYCHA) buildings subject to the policy and 11 privately managed buildings in which most residents received housing vouchers (herein "Section 8"). We invited participants from nonsmoking households (NYCHA n = 157, Section-8 n = 118) to enroll in a longitudinal air monitoring study, measuring (1) nicotine concentration with passive, bisulfate-coated filters, and (2) particulate matter (PM2.5) with low-cost particle sensors. We also measured nicotine concentrations and counted cigarette butts in common areas (n = 91 stairwells and hallways). We repeated air monitoring sessions in households and common areas every 6 months, totaling six post-policy sessions. After 3 years, we observed larger declines in nicotine concentration in NYCHA hallways than in Section-8, [difference-in-difference (DID) = -1.92 µg/m3 (95% CI -2.98, -0.87), p = .001]. In stairwells, nicotine concentration declines were larger in NYCHA buildings, but the differences were not statistically significant [DID= -1.10 µg/m3 (95% CI -2.40, 0.18), p = .089]. In households, there was no differential change in nicotine concentration (p = .093) or in PM2.5 levels (p = .385). Nicotine concentration reductions in NYCHA common areas over 3 years may be attributable to the SFH policy, reflecting its gradual implementation over this time. Continued air monitoring over multiple years has demonstrated that SHS exposure may be declining more rapidly in NYCHA common areas as a result of SFH policy adherence. This may have positive implications for improved health outcomes among those living in public housing, but additional tracking of air quality and studies of health outcomes are needed. Ongoing efforts by NYCHA to integrate the SFH policy into wider healthier-homes initiatives may increase policy compliance.

Feasibility of low-cost particle sensor types in long-term indoor air pollution health studies after repeated calibration, 2019–2021

Previous studies have explored using calibrated low-cost particulate matter (PM) sensors, but important research gaps remain regarding long-term performance and reliability. Evaluate longitudinal performance of low-cost particle sensors by measuring sensor performance changes over 2 years of use. 51 low-cost particle sensors (Airbeam 1 N = 29; Airbeam 2 N = 22) were calibrated four times over a 2-year timeframe between 2019 and 2021. Cigarette smoke-specific calibration curves for Airbeam 1 and 2 PM sensors were created by directly comparing simultaneous 1-min readings of a Thermo Scientific Personal DataRAM PDR-1500 unit with a 2.5 µm inlet. Inter-sensor variability in calibration coefficient was high, particularly in Airbeam 1 sensors at study initiation. Calibration coefficients for both sensor types trended downwards over time to < 1 at final calibration timepoint [Airbeam 1 Mean (SD) = 0.87 (0.20); Airbeam 2 Mean (SD) = 0.96 (0.27)]. We lost more Airbeam 1 sensors (N = 27 out of 56, failure rate 48.2%) than Airbeam 2 (N = 2 out of 24, failure rate 8.3%) due to electronics, battery, or data output issues. Evidence suggests degradation over time might depend more on particle sensor type, rather than individual usage. Repeated calibrations of low-cost particle sensors may increase confidence in reported PM levels in longitudinal indoor air pollution studies.

Evaluation of a new low-cost particle sensor as an internet-of-things device for outdoor air quality monitoring

ABSTRACT Many low-cost particle sensors are available for routine air quality monitoring of PM2.5, but there are concerns about the accuracy and precision of the reported data, particularly in humid conditions. The objectives of this study are to evaluate the Sensirion SPS30 particulate matter (PM) sensor against regulatory methods for measurement of real-time particulate matter concentrations and to evaluate the effectiveness of the Intelligent AirTM sensor pack for remote deployment and monitoring. To achieve this, we co-located the Intelligent AirTM sensor pack, developed at Clemson University and built around the Sensirion SPS30, to collect data from July 29, 2019, to December 12, 2019, at a regulatory site in Columbia, South Carolina. When compared to the Federal Equivalent Methods, the SPS30 showed an average bias adjusted R2 = 0.75, mean bias error of −1.59, and a root mean square error of 2.10 for 24-hour average trimmed measurements over 93 days, and R2 = 0.57, mean bias error of −1.61, and a root mean square error of 3.029, for 1-hr average trimmed measurements over 2300 hours when the central 99% of data was retained with a data completeness of 75% or greater. The Intelligent AirTM sensor pack is designed to promote long-term deployment and includes a solar panel and battery backup, protection from the elements, and the ability to upload data via a cellular network. Overall, we conclude that the SPS30 PM sensor and the Intelligent AirTM sensor pack have the potential for greatly increasing the spatial density of particulate matter measurements, but more work is needed to understand and calibrate sensor measurements. Implications: This work adds to the growing body of research that indicates that low-cost sensors of particulate matter (PM) for air quality monitoring has a promising future, and yet much work is left to be done. This work shows that the level of data processing and filtering effects how the low-cost sensors compare to existing federal reference and equivalence methods: more data filtering at low PM levels worsens the data comparison, while longer time averaging improves the measurement comparisons. Improvements must be made to how we handle, calibrate, and correct PM data from low-cost sensors before the data can be reliably used for air quality monitoring and attainment.

Differentiating Semi-Volatile and Solid Particle Events Using Low-Cost Lung-Deposited Surface Area and Black Carbon Sensors

Low-cost particle sensors have proven useful in applications such as source apportionment, health, and reactivity studies. The benefits of these instruments increase when used in parallel, as exemplified with a 3-month long deployment in an urban background site. Using two lung-deposited surface area (LDSA) instruments, a low-cost method was developed to assess the solid component of an aerosol by applying a catalytic stripper to the inlet stream of one LDSA instrument, resulting in only the solid fraction of the sample being measured (LDSAc). To determine the semi-volatile fraction of the sample, the LDSAC was compared to the LDSA without a catalytic stripper, thus measuring all particles (LDSAN). The ratio of LDSA (LDSAC/LDSAN) was used to assess the fraction of solid and semi-volatile particles within a sample. Here, a low ratio represents a high fraction of semi-volatile particles, with a high ratio indicating a high fraction of solid particles. During the 3-month urban background study in Birmingham, UK, it is shown that the LDSA ratios ranged from 0.2–0.95 indicating a wide variation in sources and subsequent semi-volatile fraction of particles. A black carbon (BC) instrument was used to provide a low-cost measure of LDSA to BC ratio. Comparatively, the LDSA to BC ratios obtained using low-cost sensors showed similar results to high-cost analyses for urban environments. During a high LDSAC/LDSAN ratio sampling period, representing high solid particle concentrations, an LDSA to BC probability distribution was shown to be multimodal, reflecting urban LDSA to BC ratio distributions measured with laboratory-grade instrumentation. Here, a low-cost approach for data analyses presents insight on particle characteristics and insight into PM composition and size, useful in source apportionment, health, and atmospheric studies.

Significance of Meteorological Feature Selection and Seasonal Variation on Performance and Calibration of a Low-Cost Particle Sensor

Poor air quality is a major environmental concern worldwide, but people living in low- and middle-income countries are disproportionately affected. Measurement of PM2.5 is essential for establishing regulatory standards and developing policy frameworks. Low-cost sensors (LCS) can construct a high spatiotemporal resolution PM2.5 network, but the calibration dependencies and subject to biases of LCS due to variable meteorological parameters limit their deployment for air-quality measurements. This study used data collected from June 2019 to April 2021 from a PurpleAir Monitor and Met One Instruments’ Model BAM 1020 as a reference instrument at Alberta, Canada. The objective of this study is to identify the relevant meteorological parameters for each season that significantly affect the performance of LCS. The meteorological features considered are relative humidity (RH), temperature (T), wind speed (WS) and wind direction (WD). This study applied Multiple Linear Regression (MLR), k-Nearest Neighbor (kNN), Random Forest (RF) and Gradient Boosting (GB) models with varying features in a stepwise manner across all the seasons, and only the best results are presented in this study. Improvement in the performance of calibration models is observed by incorporating different features for different seasons. The best performance is achieved when RF is applied but with different features for different seasons. The significant meteorological features are PM2.5_LCS in Summer, PM2.5_LCS, RH and T in Autumn, PM2.5_LCS, T and WS in Winter and PM2.5_LCS, RH, T and WS in Spring. The improvement in R2 for each season (values in parentheses) is Summer (0.66–0.94), Autumn (0.73–0.96), Winter (0.70–0.95) and Spring (0.70–0.94). This study signifies selecting the right combination of models and features to attain the best results for LCS calibration.

Intercomparison of PurpleAir Sensor Performance over Three Years Indoors and Outdoors at a Home: Bias, Precision, and Limit of Detection Using an Improved Algorithm for Calculating PM2.5

Low-cost particle sensors are now used worldwide to monitor outdoor air quality. However, they have only been in wide use for a few years. Are they reliable? Does their performance deteriorate over time? Are the algorithms for calculating PM2.5 concentrations provided by the sensor manufacturers accurate? We investigate these questions using continuous measurements of four PurpleAir monitors (8 sensors) under normal conditions inside and outside a home for 1.5–3 years. A recently developed algorithm (called ALT-CF3) is compared to the two existing algorithms (CF1 and CF_ATM) provided by the Plantower manufacturer of the PMS 5003 sensors used in PurpleAir PA-II monitors. Results. The Plantower CF1 algorithm lost 25–50% of all indoor data due in part to the practice of assigning zero to all concentrations below a threshold. None of these data were lost using the ALT-CF3 algorithm. Approximately 92% of all data showed precision better than 20% using the ALT-CF3 algorithm, but only approximately 45–75% of data achieved that level using the Plantower CF1 algorithm. The limits of detection (LODs) using the ALT-CF3 algorithm were mostly under 1 µg/m3, compared to approximately 3–10 µg/m3 using the Plantower CF1 algorithm. The percentage of observations exceeding the LOD was 53–92% for the ALT-CF3 algorithm, but only 16–44% for the Plantower CF1 algorithm. At the low indoor PM2.5 concentrations found in many homes, the Plantower algorithms appear poorly suited.

Influence of particle properties and environmental factors on the performance of typical particle monitors and low-cost particle sensors in the market of China

The performance of particle measurements based on three types of principles, including tapered element oscillating microbalance (TEOM), beta attenuation monitor (BAM) and light scattering, were investigated with different types of particle (braize, arizona dust and polystyrene spheres), particle sizes (1.1, 2.0, 2.5 and 3.0 μm), and environmental factors (temperature, relative humidity and relatively high particle concentration). Specifically, two middle-cost monitors (LY, LL) and three low-cost sensors (PT, YT, SF) with light scattering were considered.Good correlations with gravimetric were all observed on three principle instruments with three particle types at laboratory tests, except for less concerned intercepts did not meet the requirements of standard HJ 653–2013. Particle properties showed negligible effect on the instrumental measurements based on BAM and TEOM, commonly as high-cost monitors. Interestingly, middle-cost monitors were found to be more sensitive to particle properties versus low-cost sensors using calibration dusts with various properties. Moreover, obvious impacts of particle sizes were observed for all tested light-scattering monitors and sensors. In addition, different trends of sensors measuring PM10/PM2.5/PM1.0 with the increase of particle sizes indicated the poor particle recognition of light scattering principle. Expectedly, poor correlations of sensors were observed in ambient experiments compared with laboratory experiments due to the effect of complex environmental factors. Relative humidity (RH) and temperature (T) were proposed to express synergistic effects on measurements by sensors. Meanwhile, poor correlation and zero drift of YT sensors after high particle concentration exposure experiments revealed high particle concentrations would result in the pollution of low-cost sensors due to the lack of self-cleaning function. In conclusion, to evaluate the performance of sensors for future standard instrumental developments more accurately, both indoor and outdoor assessments should be taken out. The data acquired in this study could provide information for future evaluation of the performance of particle monitors and sensors based on light scattering principle.

Performance and Deployment of Low-Cost Particle Sensor Units to Monitor Biomass Burning Events and Their Application in an Educational Initiative.

Biomass burning smoke is often a significant source of airborne fine particles in regional areas where air quality monitoring is scarce. Emerging sensor technology provides opportunities to monitor air quality on a much larger geographical scale with much finer spatial resolution. It can also engage communities in the conversation around local pollution sources. The SMoke Observation Gadget (SMOG), a unit with a Plantower dust sensor PMS3003, was designed as part of a school-based Science, Technology, Engineering and Mathematics (STEM) project looking at smoke impacts in regional areas of Victoria, Australia. A smoke-specific calibration curve between the SMOG units and a standard regulatory instrument was developed using an hourly data set collected during a peat fire. The calibration curve was applied to the SMOG units during all field-based validation measurements at several locations and during different seasons. The results showed strong associations between individual SMOG units for PM2.5 concentrations (r2 = 0.93–0.99) and good accuracy (mean absolute error (MAE) < 2 μg m−3). Correlations of the SMOG units to reference instruments also demonstrated strong associations (r2 = 0.87–95) and good accuracy (MAE of 2.5–3.0 μg m−3). The PM2.5 concentrations tracked by the SMOG units had a similar response time as those measured by collocated reference instruments. Overall, the study has shown that the SMOG units provide relevant information about ambient PM2.5 concentrations in an airshed impacted predominantly by biomass burning, provided that an adequate adjustment factor is applied.

Improving data reliability: A quality control practice for low-cost PM2.5 sensor network.

Low-cost air quality sensor networks have been increasingly used for high spatial resolution air quality monitoring in recent years. Ensuring data reliability during continuous operation is critical for these sensor networks. Using particulate matter sensor as an example, this study reports a data quality control method, including sensor selection, pre-calibration, and online inspection. It was used in developing and operating the dense low-cost particle sensor networks in two Chinese cities. Firstly, seven mainstream sensors were tested and one model of particle sensor was selected due to its better linearity and stability. For a batch of sensors of the same model, although they were calibrated after manufactured, there are differences in response toward the same concentration of pollutants. The systematical variation of sensors was corrected and unified through pre-calibration. After deploying them in the field, a data analysis method is established for online inspecting their working status. Using data from these sensors, it evaluates parameters such as intraclass correlation coefficients and normalized root mean square error. These two metrics help to construct a two-dimensional coordinate system and to classify sensors into four status, including normal, fluctuation, hotspots, and malfunction. During a one-month operation in the two cities, 8 (out of 82) and 10 (out of 59) sensors with suspected malfunctions were screened out for further on-site inspection. Moreover, the sensor networks show potential in identifying illegal emission sources that cannot be typically detected by sparse regulatory air quality monitoring stations.

A systematic investigation on the effects of temperature and relative humidity on the performance of eight low-cost particle sensors and devices

Low-cost particle sensors are rapidly gaining popularity for indoor air quality monitoring, but their accuracy has been shown to vary among products and among environments in which they operate. Previous research has explored the effect of environmental conditions including temperature and humidity on sensor accuracy and found appreciable effects of these variables in many cases. We extend this work to cover some models that have not been previously tested and analyze results in a way to establish the significance of environmental variables on low-cost sensor output. In this study, we conduct a series of laboratory experiments in which temperature and relative humidity (RH) are varied in a precisely controlled chamber and different common particle sources introduced. Eight different brands of low-cost particle sensors, some of which have not been previously analyzed in this manner, measured particle concentration inside the chamber and were compared to research-grade instruments. Comparison of the output of the research-grade instruments with that of the low-cost sensors yielded a few interesting results. First, temperature showed no statistically significant effect on sensor performance in any case, at 99% confidence. Second, RH appears to affect the magnitude of low-cost sensor output but not its correlation to research-grade instruments output, suggesting the possibility of RH-based calibration.

A machine learning field calibration method for improving the performance of low-cost particle sensors

Measurements of airborne particles in buildings with low-cost optical particle counters (OPCs) are often inaccurate and subject to uncertainties. This study introduces a methodology to improve the performance of low-cost OPCs in measuring indoor particles through machine learning. A two-month field measurement campaign was conducted in an occupied net-zero energy house. The studied OPCs (OPC–N2, Alphasense Ltd.) report size fractionated concentrations from 0.38 to 17.5 μm. Co-located reference instrumentation included a scanning mobility particle sizer (SMPS: 0.01–0.30 μm) and an optical particle sizer (OPS: 0.30–10 μm). The machine learning field calibration method applies Gaussian Process Regression (GPR) and includes two components: (1.) correction of the size-resolved OPC counting efficiency from 0.38 to 10 μm and (2.) prediction of volume size distributions (mass proxy) below the 0.38 μm detection limit of the OPC. The field calibration method is applicable to OPCs that report size fractionated concentrations. In (1.), a GPR function was used to correct the size-resolved counting efficiency of the OPCs between 0.38 and 10 μm using the OPS as reference. In (2.), a second GPR function was used to predict the volume size distribution below 0.38 μm using the SMPS/OPS as reference. This was done given the significant contribution of sub-0.38 μm particles to volume concentrations in the accumulation mode. The machine learning field calibration method resulted in a significant improvement in the accuracy of size-integrated volume concentrations (PV2.5, PV10) reported by the OPCs as compared to the SMPS/OPS. Improvements were seen in the Pearson coefficient (before correction: 0.59–0.83; after correction: 0.98–0.99); coefficient of determination (before correction: 0.35–0.69; after correction: 0.97–0.98); and mean absolute percentage error (before correction: 35–69%; after correction: 19–25%).