Leveraging low-cost sensors to predict nitrogen dioxide for epidemiologic exposure assessment.

BACKGROUND: Statistical models of air pollution enable intra-urban characterization of pollutant concentrations, benefiting exposure assessment for environmental epidemiology. The new generation of low-cost sensors facilitate the deployment of dense monitoring networks and can potentially be used to improve intra-urban models of air pollution. OBJECTIVE: Develop and evaluate a spatiotemporal model for nitrogen dioxide (NO2) in the Puget Sound region of WA, USA for the Adult Changes in Thought Air Pollution (ACT-AP) study and assess the contribution of low-cost sensor data to the model's performance through cross-validation. METHODS: We developed a spatiotemporal NO2 model for the study region incorporating data from 11 agency locations, 364 supplementary monitoring locations, and 117 low-cost sensor (LCS) locations for the 1996-2020 time period. Model features included long-term time trends and dimension-reduced land use regression. We evaluated the contribution of LCS network data by comparing models fit with and without sensor data using cross-validated (CV) summary performance statistics. RESULTS: The best performing model had one time trend and geographic covariates summarized into three partial least squares components. The model, fit with LCS data, performed as well as other recent studies (agency cross-validation: CV- root mean square error (RMSE) = 2.5 ppb NO2; CV- coefficient of determination ( R 2 ) = 0.85). Predictions of NO2 concentrations developed with LCS were higher at residential locations compared to a model without LCS, especially in recent years. While LCS did not provide a strong performance gain at agency sites (CV-RMSE = 2.8 ppb NO2; CV- R 2 = 0.82 without LCS), at residential locations, the improvement was substantial, with RMSE = 3.8 ppb NO2 and R 2 = 0.08 (without LCS), compared to CV-RMSE = 2.8 ppb NO2 and CV- R 2 = 0.51 (with LCS). IMPACT: We developed a spatiotemporal model for nitrogen dioxide (NO2) pollution in Washington's Puget Sound region for epidemiologic exposure assessment for the Adult Changes in Thought Air Pollution study. We examined the impact of including low-cost sensor data in the NO2 model and found the additional spatial information the sensors provided predicted NO2 concentrations that were higher than without low-cost sensors, particularly in recent years. We did not observe a clear, substantial improvement in cross-validation performance over a similar model fit without low-cost sensor data; however, the prediction improvement with low-cost sensors at residential locations was substantial. The performance gains from low-cost sensors may have been attenuated due to spatial information provided by other supplementary monitoring data.

Prevalence and predictors of colon and prostate cancer screening among volunteer firefighters: The United States Firefighter Cancer Assessment and Prevention Study.

BACKGROUND: Although firefighters have increased risk for colon and prostate cancer, limited information exists on screening practices for these cancers in volunteer firefighters who compose two-thirds of the US fire service. We estimated the prevalence of colon and prostate cancer screening among volunteer firefighters using eligibility criteria from 4 evidence-based screening recommendations and evaluated factors influencing screening. METHODS: We evaluated colon (n = 569) and prostate (n = 498) cancer screening prevalence in a sample of US volunteer firefighters using eligibility criteria from the US Preventive Services Taskforce (USPSTF), National Fire Protection Association, American Cancer Society, and National Comprehensive Cancer Network. We assessed associations with fire service experience, demographics, and cancer risk perception based on USPSTF guidelines. RESULTS: For those eligible based on USPSTF guidelines, colon and prostate cancer screening prevalence was 51.7% (95% CI: 45.7, 57.8) and 48.8% (95% CI: 40.0, 57.6), respectively. Higher odds of colon and prostate cancer screening were observed with older age and with some college education compared to those with less education. Fire service experience and cancer risk perception were not associated with screening practices. CONCLUSION: This is the first large study to assess colon and prostate cancer screening among US volunteer firefighters based on different screening guidelines. Our findings suggest gaps in cancer prevention efforts in the US volunteer fire service. Promoting cancer screening education and opportunities for volunteer firefighters by their fire departments, healthcare professionals, and public health practitioners, may help to address the gaps.

Wildfire Smoke Monitoring for Agricultural Safety and Health in Rural Washington.

OBJECTIVES: This study aimed to evaluate the performance of a low-cost smoke sampling platform relative to environmental and occupational exposure monitoring methods in a rural agricultural region in central Washington state. METHODS: We co-located the Thingy AQ sampling platform alongside cyclone-based gravimetric samplers, a nephelometer, and an environmental beta attenuation mass (E-BAM) monitor during August and September of 2020. Ambient particulate matter concentrations were collected during a smoke and non-smoke period and measurements were compared across sampling methods. RESULTS: We found reasonable agreement between observations from two particle sensors within the Thingy AQ platform and the nephelometer and E-BAM measurements throughout the study period, though the measurement range of the sensors was greater during the smoke period compared to the non-smoke period. Occupational gravimetric sampling methods did not correlate with PM2.5 data collected during smoke periods, likely due to their capture of larger particle sizes than those typically measured by PM2.5 ambient air quality instruments during wildfire events. CONCLUSION: Data collected before and during an intense wildfire smoke episode in September 2020 indicated that the low-cost smoke sampling platform provides a strategy to increase access to real-time air quality information in rural areas where regulatory monitoring networks are sparse if sensor performance characteristics under wildfire smoke conditions are understood. Improving access to spatially resolved air quality information could help agricultural employers protect both worker and crop health as wildfire smoke exposure increases due to the impacts of climate change. Such information can also assist employers with meeting new workplace wildfire smoke health and safety rules.

Impact of Mobile Monitoring Network Design on Air Pollution Exposure Assessment Models.

Short-term mobile monitoring campaigns are increasingly used to assess long-term air pollution exposure in epidemiology. Little is known about how monitoring network design features, including the number of stops and sampling temporality, impacts exposure assessment models. We address this gap by leveraging an extensive mobile monitoring campaign conducted in the greater Seattle area over the course of a year during all days of the week and most hours. The campaign measured total particle number concentration (PNC; sheds light on ultrafine particulate (UFP) number concentration), black carbon (BC), nitrogen dioxide (NO2), fine particulate matter (PM2.5), and carbon dioxide (CO2). In Monte Carlo sampling of 7327 total stops (278 sites × 26 visits each), we restricted the number of sites and visits used to estimate annual averages. Predictions from the all-data campaign performed well, with cross-validated R2s of 0.51-0.77. We found similar model performances (85% of the all-data campaign R2) with ∼1000 to 3000 randomly selected stops for NO2, PNC, and BC, and ∼4000 to 5000 stops for PM2.5 and CO2. Campaigns with additional temporal restrictions (e.g., business hours, rush hours, weekdays, or fewer seasons) had reduced model performances and different spatial surfaces. Mobile monitoring campaigns wanting to assess long-term exposure should carefully consider their monitoring designs.

Design and evaluation of short-term monitoring campaigns for long-term air pollution exposure assessment.

BACKGROUND: Short-term mobile monitoring campaigns to estimate long-term air pollution levels are becoming increasingly common. Still, many campaigns have not conducted temporally-balanced sampling, and few have looked at the implications of such study designs for epidemiologic exposure assessment. OBJECTIVE: We carried out a simulation study using fixed-site air quality monitors to better understand how different short-term monitoring designs impact the resulting exposure surfaces. METHODS: We used Monte Carlo resampling to simulate three archetypal short-term monitoring sampling designs using oxides of nitrogen (NOx) monitoring data from 69 regulatory sites in California: a year-around Balanced Design that sampled during all seasons of the year, days of the week, and all or various hours of the day; a temporally reduced Rush Hours Design; and a temporally reduced Business Hours Design. We evaluated the performance of each design's land use regression prediction model. RESULTS: The Balanced Design consistently yielded the most accurate annual averages; while the reduced Rush Hours and Business Hours Designs generally produced more biased results. SIGNIFICANCE: A temporally-balanced sampling design is crucial for short-term campaigns such as mobile monitoring aiming to assess long-term exposure in epidemiologic cohorts. IMPACT STATEMENT: Short-term monitoring campaigns to assess long-term air pollution trends are increasingly common, though they rarely conduct temporally balanced sampling. We show that this approach produces biased annual average exposure estimates that can be improved by collecting temporally-balanced samples.

Assessing the value of complex refractive index and particle density for calibration of low-cost particle matter sensor for size-resolved particle count and PM2.5 measurements.

Low-cost optical scattering particulate matter (PM) sensors report total or size-specific particle counts and mass concentrations. The PM concentration and size are estimated by the original equipment manufacturer (OEM) proprietary algorithms, which have inherent limitations since particle scattering depends on particles' properties such as size, shape, and complex index of refraction (CRI) as well as environmental parameters such as temperature and relative humidity (RH). As low-cost PM sensors are not able to resolve individual particles, there is a need to characterize and calibrate sensors' performance under a controlled environment. Here, we present improved calibration algorithms for Plantower PMS A003 sensor for mass indices and size-resolved number concentration. An aerosol chamber experimental protocol was used to evaluate sensor-to-sensor data reproducibility. The calibration was performed using four polydisperse test aerosols. The particle size distribution OEM calibration for PMS A003 sensor did not agree with the reference single particle sizer measurements. For the number concentration calibration, the linear model without adjusting for the aerosol properties and environmental conditions yields an absolute error (NMAE) of ~ 4.0% compared to the reference instrument. The calibration models adjusted for particle CRI and density account for non-linearity in the OEM's mass concentrations estimates with NMAE within 5.0%. The calibration algorithms developed in this study can be used in indoor air quality monitoring, occupational/industrial exposure assessments, or near-source monitoring scenarios where field calibration might be challenging.

Deployment, Calibration, and Cross-Validation of Low-Cost Electrochemical Sensors for Carbon Monoxide, Nitrogen Oxides, and Ozone for an Epidemiological Study.

We designed and built a network of monitors for ambient air pollution equipped with low-cost gas sensors to be used to supplement regulatory agency monitoring for exposure assessment within a large epidemiological study. This paper describes the development of a series of hourly and daily field calibration models for Alphasense sensors for carbon monoxide (CO; CO-B4), nitric oxide (NO; NO-B4), nitrogen dioxide (NO2; NO2-B43F), and oxidizing gases (OX-B431)-which refers to ozone (O3) and NO2. The monitor network was deployed in the Puget Sound region of Washington, USA, from May 2017 to March 2019. Monitors were rotated throughout the region, including at two Puget Sound Clean Air Agency monitoring sites for calibration purposes, and over 100 residences, including the homes of epidemiological study participants, with the goal of improving long-term pollutant exposure predictions at participant locations. Calibration models improved when accounting for individual sensor performance, ambient temperature and humidity, and concentrations of co-pollutants as measured by other low-cost sensors in the monitors. Predictions from the final daily models for CO and NO performed the best considering agreement with regulatory monitors in cross-validated root-mean-square error (RMSE) and R2 measures (CO: RMSE = 18 ppb, R2 = 0.97; NO: RMSE = 2 ppb, R2 = 0.97). Performance measures for NO2 and O3 were somewhat lower (NO2: RMSE = 3 ppb, R2 = 0.79; O3: RMSE = 4 ppb, R2 = 0.81). These high levels of calibration performance add confidence that low-cost sensor measurements collected at the homes of epidemiological study participants can be integrated into spatiotemporal models of pollutant concentrations, improving exposure assessment for epidemiological inference.

Health Impact Assessment of the 2020 Washington State Wildfire Smoke Episode: Excess Health Burden Attributable to Increased PM2.5 Exposures and Potential Exposure Reductions.

Major wildfires starting in the summer of 2020 along the west coast of the United States made PM2.5 concentrations in this region rank among the highest in the world. Washington was impacted both by active wildfires in the state and aged wood smoke transported from fires in Oregon and California. This study aims to estimate the magnitude and disproportionate spatial impacts of increased PM2.5 concentrations attributable to these wildfires on population health. Daily PM2.5 concentrations for each county before and during the 2020 Washington wildfire episode (September 7-19) were obtained from regulatory air monitors. Utilizing previously established concentration-response function (CRF) of PM2.5 (CRF of total PM2.5) and odds ratio (OR) of wildfire smoke days (OR of wildfire smoke days) for mortality, we estimated excess mortality attributable to the increased PM2.5 concentrations in Washington. On average, daily PM2.5 concentrations increased 97.1 µg/m3 during the wildfire smoke episode. With CRF of total PM2.5, the 13-day exposure to wildfire smoke was estimated to lead to 92.2 (95% CI: 0.0, 178.7) more all-cause mortality cases; with OR of wildfire smoke days, 38.4 (95% CI: 0.0, 93.3) increased all-cause mortality cases and 15.1 (95% CI: 0.0, 27.9) increased respiratory mortality cases were attributable to the wildfire smoke episode. The potential impact of avoiding elevated PM2.5 exposures during wildfire events significantly reduced the mortality burden. Because wildfire smoke episodes are likely to impact the Pacific Northwest in future years, continued preparedness and mitigations to reduce exposures to wildfire smoke are necessary to avoid excess health burden.

Use of low-cost PM monitors and a multi-wavelength aethalometer to characterize PM2.5 in the Yakama Nation Reservation.

Rural lower Yakima Valley, Washington is home to the reservation of the Confederated Tribes and Bands of the Yakama Nation, and is a major agricultural region. Episodic poor air quality impacts this area, reflecting sources of particulate matter with a diameter of less than 2.5 micrometers (PM2.5) that include residential wood smoke, agricultural biomass burning and other emissions, truck traffic, backyard burning, and wildfire smoke. University of Washington partnered with the Yakama Nation Environmental Management Program to investigate characteristics of PM2.5 using 9 months of data from a combination of low-cost optical particle counters and a 5-wavelength aethalometer (MA200 Aethlabs) over 4 seasons and an episode of summer wildfire smoke. The greatest percentage of hours sampled with PM2.5 >12 µg/m3 occurred during the wildfire smoke episode (59%), followed by fall (23%) and then winter (21%). Mean (SD) values of Delta-C (µg/m3), which has been posited as an indicator of wood smoke, and determined as the mass absorbance difference at 375-880nm, were: summer - wildfire smoke 0.34 (0.52), winter 0.27 (0.32), fall 0.10 (0.22), spring 0.05 (0.11), and summer - no wildfire smoke 0.04 (0.14). Mean (95% confidence interval) values of the absorption Ångström exponent, an indicator of the wavelength dependence of the aerosol, were: winter 1.5 (1.2-1.8), summer - wildfire smoke 1.4 (1.0-1.8), fall 1.3 (1.1-1.4), spring 1.2 (1.1-1.4), and summer - no wildfire smoke 1.2 (1.0-1.3). The trends in Delta-C and absorption Ångström exponents are consistent with expectations that a higher value reflects more biomass burning. These results suggest that biomass burning is an important contributor to PM2.5 in the wintertime, and emissions associated with diesel and soot are important contributors in the fall; however, the variety of emissions sources and combustion conditions present in this region may limit the utility of traditional interpretations of aethalometer data. Further understanding of how to interpret aethalometer data in regions with complex emissions would contribute to much-needed research in communities impacted by air pollution from agricultural as well as residential sources of combustion.

Health Impact Assessment of PM2.5 attributable mortality from the September 2020 Washington State Wildfire Smoke Episode.

Major wildfires that started in the summer of 2020 along the west coast of the U.S. have made PM2.5 concentrations in cities in this region rank among the highest in the world. Regions of Washington were impacted by active wildfires in the state, and by aged wood smoke transported from fires in Oregon and California. This study aims to assess the population health impact of increased PM2.5 concentrations attributable to the wildfire. Average daily PM2.5 concentrations for each county before and during the 2020 Washington wildfire episode were obtained from the Washington Department of Ecology. Utilizing previously established associations of short-term mortality for PM2.5, we estimated excess mortality for Washington attributable to the increased PM2.5 levels. On average, PM2.5 concentrations increased 91.7 µg/m3 during the wildfire episode. Each week of wildfire smoke exposures was estimated to result in 87.6 (95% CI: 70.9, 103.1) cases of increased all-cause mortality, 19.1 (95% CI: 10.0, 28.2) increased cardiovascular disease deaths, and 9.4 (95% CI: 5.1, 13.5) increased respiratory disease deaths. Because wildfire smoke episodes are likely to continue impacting the Pacific Northwest in future years, continued preparedness and mitigations to reduce exposures to wildfire smoke are necessary to avoid this excess health burden.

Calibration of low-cost particulate matter sensors: Model development for a multi-city epidemiological study.

Low-cost air monitoring sensors are an appealing tool for assessing pollutants in environmental studies. Portable low-cost sensors hold promise to expand temporal and spatial coverage of air quality information. However, researchers have reported challenges in these sensors' operational quality. We evaluated the performance characteristics of two widely used sensors, the Plantower PMS A003 and Shinyei PPD42NS, for measuring fine particulate matter compared to reference methods, and developed regional calibration models for the Los Angeles, Chicago, New York, Baltimore, Minneapolis-St. Paul, Winston-Salem and Seattle metropolitan areas. Duplicate Plantower PMS A003 sensors demonstrated a high level of precision (averaged Pearson's r = 0.99), and compared with regulatory instruments, showed good accuracy (cross-validated R2 = 0.96, RMSE = 1.15 µg/m3 for daily averaged PM2.5 estimates in the Seattle region). Shinyei PPD42NS sensor results had lower precision (Pearson's r = 0.84) and accuracy (cross-validated R2 = 0.40, RMSE = 4.49 µg/m3). Region-specific Plantower PMS A003 models, calibrated with regulatory instruments and adjusted for temperature and relative humidity, demonstrated acceptable performance metrics for daily average measurements in the other six regions (R2 = 0.74-0.95, RMSE = 2.46-0.84 µg/m3). Applying the Seattle model to the other regions resulted in decreased performance (R2 = 0.67-0.84, RMSE = 3.41-1.67 µg/m3), likely due to differences in meteorological conditions and particle sources. We describean approach to metropolitan region-specific calibration models for low-cost sensors that can be used with cautionfor exposure measurement in epidemiological studies.

Laboratory Evaluation of the Shinyei PPD42NS Low-Cost Particulate Matter Sensor.

OBJECTIVE: Finely resolved PM2.5 exposure measurements at the level of individual participants or over a targeted geographic area can be challenging due to the cost, size and weight of the monitoring equipment. We propose re-purposing the low-cost, portable and lightweight Shinyei PPD42NS particle counter as a particle counting device. Previous field deployment of this sensor suggests that it captures trends in ambient PM2.5 concentrations, but important characteristics of the sensor response have yet to be determined. Laboratory testing was undertaken in order to characterize performance. METHODS: The Shinyei sensors, in-line with a TSI Aerosol Particle Sizer (APS) model 3321, tracked particle decay within an aerosol exposure chamber. Test atmospheres were composed of monodisperse polystyrene spheres with diameters of 0.75, 1, 2 3 and 6 um as well as a polydisperse atmosphere of ASHRAE test dust #1. RESULTS: Two-minute block averages of the sensor response provide a measurement with low random error, within sensor, for particles in the 0.75-6µm range with a limit of detection of 1 µg/m3. The response slope of the sensors is idiomatic, and each sensor requires a unique response curve. A linear model captures the sensor response for concentrations below 50 µg/m3 and for concentrations above 50 µg/m3 a non-linear function captures the response and saturates at 800 µg/m3. The Limit of Detection (LOD) is 1 µg/m3. The response time is on the order of minutes, making it appropriate for tracking short-term changes in concentration. CONCLUSIONS: When paired with prior evaluation, these sensors are appropriate for use as ambient particle counters for low and medium concentrations of respirable particles (< 100 ug/m3). Multiple sensors deployed over a spatial grid would provide valuable spatio-temporal variability in PM2.5 and could be used to validate exposure models. When paired with GPS tracking, these devices have the potential to provide time and space resolved exposure measurements for a large number of participants, thus increasing the power of a study.