Poster Presentations

We are excited to host over 85 poster presentations for this year's ASIC in Pasadena, California. The posters will be displayed in the Pasadena Convention Center's exhibit hall for the full duration of the conference to allow for ample time to review. We encourage all in-person attendees to visit the poster displays, grouped by session during the multiple breaks throughout the week.

These posters will also be displayed on the virtual platform for all our virtual and in-person attendees to review and ask questions about. The posters will be displayed as a PDF on the platform along with an additional 1-2 minute video by the presenter giving a brief overview of the poster topic.

Poster presenters will be able to have one on one conversations about their poster in-person in the exhibit hall during the Welcome Reception & Poster Review or virtually on-line during the Thursday morning virtual activity session. More details to come!

Skip to the posters on each topic by clicking below:

Air Sensor Use in India

Practical Lessons for Village Sensor Networks from South India

Presented By: Ajay Pillarisetti, University of Pennsylvania

The motivations driving efforts to deploy lower-cost sensor (LCS) networks in urban areas apply equally to the far less-studied rural regions of low- and middle-income countries (LMICs). Here, several lines of evidence indicate that air pollution sources at household, village, and regional scales can combine to generate air quality as unhealthy as in nearby towns or cities. Yet ground-level empirical data, which would expand and deepen comprehension of exposures, sources, and solutions, remain sparse, incommensurate with the potential magnitude of the public health risk. In order to help close this gap, our research group is currently deploying LCS networks across villages in central Tamil Nadu for the AAM-LASSI project. As a preparatory step, we piloted two dozen devices for a three month interval among such communities. We encountered overlapping challenges distinct to LMIC rural settings. Communicating the lessons learned is the focus of this presentation. The practical considerations can be distilled into three categories: (1) device specification, to compensate, primarily, for less reliable mobile connectivity and mains power; (2) calibration strategies, to efficiently correct raw data by collocation with a reference-grade instrument; and (3) community engagement, to facilitate informed cooperation with the local populace. Addressing these issues has refined our approach to the main LCS network study. Furthermore, though our emphasis is academic research on air pollution in the Indian subcontinent, we anticipate that many of the points discussed will be relevant to similar projects and for other purposes, such as regulatory monitoring and citizen science, within resource-poor rural areas. We encourage such endeavors to share their experiences with the ultimate goal of fostering the iterative development of best practices for deploying LCS networks in rural regions of LMICs.

Additional Authors: Manish Desai, Adithi Upadhya, Ajay Pillarisetti, Arulselvan Sadasivam, Durairaj Natesan, Gurusamy Thangavel, Junaid Khader, Kalpana Balakrishnan, Krishnendu Mukhopadhyay, Meenakchi Sundarem Gopal Krishna, Naveen Puttaswamy, Omprashanth Rajan, Priyakumar Natarajan, Rengaraj Ramasamy, Ronak Sutaria, Sankar Sambandam, Saritha Sendhil, Sreekanth Vakacherla, Srinivasan Natarajan, Sudhakar Rao, Sylesh Loganathan

In-Person Poster Display: #1

Clean Air Monitoring and Solutions Network: getting useful, actionable data out of low cost sensors for air quality action

Network Calibration and Wildfire Data Correction for Low-Cost Air Quality Sensors: Lessons Learned from the Richmond Air Monitoring Network

Presented By: Audrey Smith Physicians, Scientists and Engineers for Healthy Energy (PSE)

Low-cost air monitors show great promise for improving the spatiotemporal resolution of air quality monitoring data, with implications for regulatory, academic, and citizen science contexts. Before the full potential of this technology is met, however, significant data management challenges must be addressed. Robust quality control procedures are needed to discern between valid spatiotemporal variability in air pollutant concentrations and artifacts of sensor drift, sensitivity to ambient conditions, and inherent inter-device variability. Moreover, although air monitoring is of particular interest during wildfire events, low-cost sensor performance may be sub-optimal in the presence of wildfire smoke, necessitating additional quality control activities. We enacted a suite of data management protocols to address these concerns for 50 Aeroqual AQY units measuring O3, NO2, and PM2.5throughout Richmond and San Pablo in the East San Francisco Bay Area. We used a mean-variance moment matching approach to detect and correct for sensor drift by comparing the running probability distribution of each sensor to that of a nearby regulatory grade instrument. Using two devices co-located with a regulatory site, we found that this approach worked satisfactorily—except during wildfires, when additional corrections were required. We co-located an Aerosol Black Carbon Detector (ABCD) alongside each AQY unit for three weeks, including 12 days when wildfire smoke impacts were observed in the Bay Area. We used measured black carbon concentrations and an anomaly detection algorithm to detect ground-level smoke events on an hourly basis. Where wildfire smoke events were identified, we applied a wildfire correction to the data. Together, these activities correct for noise in the data and enable subsequent analyses to inform air pollution control efforts in the Richmond-San Pablo Area. This collection of data management protocols provides a framework for future air monitoring networks. *Selected for a Lighting Talk

Additional Authors: Audrey Smith, Rebecca Sugrue, Karan Shetty, James Butler, Chelsea Preble, Thomas Kirchstetter, Boris Lukanov

In-Person Poster Display: #23

Air Quality Sensors for Smart City Applications in the Netherlands

Presented By: Burcu Zijlstra, OnePlanet Research Center

Despite significant progress in the past decade, ambient air pollution remains a concern in the European Union, where elevated levels of airborne particulate matter and gases lead to both adverse health effects to the public and their deposition results in long-term harm to sensitive ecosystems. Highly accurate but sparsely distributed air quality measurement stations deployed by public health institutes are essential to measurement of average air quality over the country. However, a significant data gap remains: local pollution sources in cities related to human activity. To address this gap, we present a study that aims to quantify the impact of (1) two types of construction (highway and housing) in the cities of Dordrecht and Utrecht, (2) different traffic scenarios in the city of Apeldoorn in the Netherlands on air quality. IoT sensor units supplied by Connected Worlds that measure particulate matter (PM), NO2 and microclimate are spatially distributed around the pollution sources. Potential pollution-emitting activities are documented to facilitate the data analysis. Sensor data is combined with weather data to quantify the local pollution and measure the spread. We use advanced calibration algorithms to minimize the effects of local microenvironment on the low-cost sensor response and accuracy, and use sensors co-located with reference stations to evaluate raw and calibrated sensor performance over the deployment period. First results show that there is short-lasting but significant increase in particulate matter concentration during a housing demolition in Utrecht despite the undertaken pollution-reducing measures. Insights from this study can be used to both assess the suitability of IoT sensor use in measuring local pollution, and evaluate the effectiveness of pollution-reducing measures such as using water sprays to limit PM release in construction, or using a car-priority traffic scenario to reduce the number of car stops in the city.

Additional Authors: Burcu Zijlstra, Jelle Hofman, Jasper Fabius, Sharada P. Shantaram, Hans Nouwens, Jeroen Keppels, Aad Vredenbregt, Valerio Panzica La Manna

In-Person Poster Display: #6

Applicability of factory calibrated optical particle counters for high-density air quality monitoring networks in Ghana

Presented By: Collins Gameli Hodoli, Clean Air One Atmosphere

The utility of low-cost sensors (LCS) provides new opportunities for advancing air quality monitoring in Africa. Of a particular interest, the performance and robustness of LCS under tropical conditions and highly polluted settings is currently being investigated. As a proof-of-concept, this study aimed to firstly test the functionality of factory calibrated optical particle counters (OPC) particularly Alphasense OPC-N2 for reporting meaningful air quality data in a complex environment of varying atmospheric emission sources. Secondly, whether the reported LCS data will be useful for understanding daily trends and sources of atmospheric emissions using open-source air pollution data analytical tool “openair” package. Two multi-sensor nodes manufactured by Atmospheric Sensors Ltd (ASL) were used for this study which were capable of monitoring PM (PM10, PM2.5 and emerging PM1); VOCs, NOx, CO, CO2 and O3. Only the OPC-N2 data was used in this study. On reproducibility, Pearson’s correlation analysis (r = 0.97 and 0.98) revealed that the OPC-N2 is capable of reporting meaningful PM data to support local air quality monitoring campaigns. Trend and source feature extraction analysis using modelled wind component data showed that the reported OPC-N2 data is a useful tool for understanding routine air quality levels and emission sources. Daily PM levels ranges from 500µ/m3, 90µ/m3 and 60 µ/m3 for PM10, PM2.5 and PM1 respectively. PM sources were local at wind speeds ≤ 2ms-1linked to background anthropogenic activities. These findings agree with previous works and current findings on PM pollution in Ghana and similar environments which have demonstrated the feasibility of the OPC-N2 for bridging the huge data gap and to complement sporadic air quality monitoring regimes in Ghana. This benchmark study is to support air pollution research and to help track mitigation strategies by governmental agencies in Ghana and similar environments using LCS. *Selected for a Lighting Talk

Additional Authors: Collins Gameli Hodoli, Frederic Coulon, Iq Mead

In-Person Poster Display: #8

Community Awareness and Risk Perception of Industrial Air Pollution in Rural Kenya

Presented By: Eunice Omanga, URADCA

Background: Developing countries have limited air quality management systems due to inadequate legislation and lack of political will among others. Maintaining a balance between economic development and sustainable environment is a challenge hence investments in pollution prevention technologies get sidelined in favor of short-term benefits from increased production and job creation. This lack of air quality management capability translates into lack of air pollution data, hence the false belief that there is no problem. Objectives: Assess the population’s environmental awareness, explore their perception of air pollution threat to their health; and identify the most important factors influencing their perception. Methods: A quantitative questionnaire gathered information on demographic, health status, environmental perception and environmental knowledge of residents to understand their view of pollution in their neighborhood. Focus group discussions (FGDs) allowed for corroboration of the quantitative data. Results: Four out of five respondents perceived industrial air pollution as posing a considerable risk to them despite the fact that the industry was the largest employer in the area. Respondents also argued that they had not being actively involved in identifying solutions to the environmental challenges. The study revealed the most important factors influencing the residents’ air pollution risk perception were environmental awareness and family health status. Conclusion: This study availed information to policy makers and researchers concerning public awareness and attitudes towards environmental air pollution pertinent to development and implementation of environmental policies for public health. Key words: Environmental, perception, air, pollution, risk, rural. *Selected for a Lighting Talk

Additional Authors: Eunice Omanga

In-Person Poster Display: #9

Bi-weekly low-cost NO2 sensor collocation for improved calibration performance

Presented By: Jason A. Miech, Arizona State University

The calibration of low-cost air quality sensors (LCSs) has been of great interest lately as their capabilities increase and the need to supplement aging air monitoring networks increases. For LCSs measuring nitrogen dioxide (NO2) the calibration models need to account for several environmental factors, such as relative humidity, temperature, and interfering species (O3). In this study, we examined the impact of extreme, fluctuating environmental conditions on the performance of twelve LCS NO2 sensors in Maricopa County, AZ from April 2021 to October 2021 at twelve different O3 monitoring sites, three of which also had a NO2 Federal Reference Method (FRM) instrument. We found that changing temperature and RH directly impacted calibration performance. When comparing the initial 2-week calibration period in April (T= 25°C, RH= 20%) to a collocation period in July (T= 34°C, RH= 44%), the root-mean-square error (RMSE) increased by an average of 74% (∆2.4 ppb) and the R2 decreased by 97% (∆0.8). To address this issue, we regularly rotated the sensors between the nine O3 sites and the three NO2 sites and used data from these collocations to develop several calibration models for each sensor unique to that period and its environmental conditions. Additionally, we randomized the dates to be used in the training and evaluation datasets to ensure an even distribution of the environmental factors. Using this period-specific calibration method, the RMSE of the April period compared to the July period increased by an average of 7% (∆0.4 ppb) and the R2decreased by 44% (∆0.4), a substantial improvement over a calibration solely based on the initial collocation. The primary conclusion is that NO2 LCS calibration performance sharply drops off when they encounter environmental conditions not included in their calibration training. We will show how focusing the calibration training dataset on more period-specific conditions can greatly improve the LCS performance. *Selected for a Lighting Talk

Additional Authors: Jason A. Miech, Levi Stanton, Meiling Gao, Paolo Micalizzi, Joshua Uebelherr, Pierre Herckes, Matthew P. Fraser

In-Person Poster Display: #10

Real-time combustion emissions monitoring in Mexicali using networked sensors

Presented By: Julien Caubel, Distributed Sensing Technologies

Mexicali is a community on the California-Mexico border that hosts a large border crossing, several highways, and many industrial/shipping facilities. Furthermore, local residential heating and cooking needs are still often met using wood and other biomass fuels. As a result, residents are disproportionately affected by air pollution from transportation, residential heaters, and other combustion sources harmful to human health. Despite these impacts, air quality (AQ) monitoring in Mexicali remains limited (~5 stations cover the 44 sq. mile municipality), as existing AQ instruments are typically expensive and difficult to deploy. We operated ObservAir sensors at two regulatory sites in Mexicali to monitor concentrations of black carbon (BC) aerosols, carbon monoxide, and nitrogen oxide. These pollutants are products of incomplete combustion, and are therefore strongly correlated to the combustion activities that drive harmful air quality. The sensors are also outfitted with wireless communications, patented environmental compensation features, and a weatherproof enclosure that enable long-term, outdoor deployments nearly anywhere. Throughout the study, sensors were collocated with regulatory PM2.5and BC monitors, and data was sent in real-time to a dedicated web dashboard that provides pollution alerts and maintenance notifications. Using the data collected, we investigate AQ trends and validate the ObservAir’s BC data against the regulatory standard. The dataset demonstrates that the ObservAir accurately monitors combustion-specific pollutants, providing a more comprehensive picture of AQ impacts than PM2.5measurements alone. With sufficient context, the multi-pollutant data may be used to attribute pollution contributions from different combustion sources. When analyzed alongside relevant health metrics, these near real-time AQ assessments will ultimately inform and validate public policies to better protect vulnerable urban populations in Mexicali and elsewhere. *Selected for a Lighting Talk

Additional Authors: Julien Caubel, Jose Landeros, Troy Cados, Sergio Valenzuela, Eva Luu, Paul Schafer

In-Person Poster Display: #13

Developing regional low-cost sensor (LCS) calibration models during wildfire episodes to improve sensor performance over broad concentration ranges

Presented By: Sakshi Jain, The University of British Columbia

It has been established previously that low-cost sensors (LCS) have environmental and cross-sensitivities that require robust calibration at regular intervals across the range of expected concentrations. However, this presents a challenge to calibrate LCS for extreme concentrations during wildfire episodes which has higher health and environmental consequences, despite being infrequent and short-term. To address this challenge, we deployed 16 LCS units to measure PM2.5 and NOx across Metro Vancouver (MV) during, before and after the 2020 wildfire episode and collected pollutant data from 6 regulatory monitoring stations. Collected RAMP data were then down-averaged to 5-min resolution and missing data points were imputed using Kalman filter. Since forest fire concentrations are regional, we used a baseline detection algorithm (rolling ball) to separate the regional component and constructed separate models for the regional (baseline) and local signals. Fitting calibrations to the regional signal removed the need for side-by-side colocation. We built a general calibration model for regional signal using median baseline concentrations across all outdoor RAMPs at each time stamp and training against median baseline signal across all MV stations using either regression or hybrid regression-random forests. A generalized calibration model was preferred over individual RAMP calibrations to make it transferable to units that may not have been deployed outdoors during wildfires. For the local signal calibration, we use collocation data before and after forest fire and calibrated via standard published approaches. Outputs from regional and local signal calibration were combined to establish the final calibrated PM2.5 or NOx during the wildfire episode. Performance will be assessed on a withheld testing data set in Vancouver. *Selected for a Lighting Talk

Additional Authors: Sakshi Jain, Melanie MacArthur, Mrinmoy Chakraborty, Surya Dhulipala, Naomi Zimmerman

In-Person Poster Display: #5

Developing a Novel Sensor Technology for Measuring Particulate Matter on Unmanned Aircrafts

Presented By: Andres Munevar, Embry-Riddle Aeronautical University

Atmospheric Particulate Matter (PM) has been identified as a detrimental criteria pollutant to human health and the environment. Current PM monitoring strategies are limited in capturing variations in atmospheric PM concentrations at fine temporal and spatial scales due to their high cost and size limitations. Recently, low-cost PM sensors (LCPMS) have been used to address these shortcomings enabling expanded coverage and higher spatiotemporal resolution. However, the lack of three-dimensional monitoring hinders our ability to fully characterize variations of PM concentrations with altitude. The aim of this study is to develop a robust method to integrate lightweight and compactLCPMS to detect PM concentrations using unmanned aircraft (UA). Three replicates of two types of optical particle counters were used namely Alphasense OPC N3 and PurpleAir PA-II-SD. To evaluate the performance of LCPMS at different meteorological conditions, calibration of the sensors was conducted in Orlando, FL for two months in the summer and winter seasons in 2021. The LCPMS were collocated with a Vaisala AQT400 PM monitor and a PM reference monitor with simultaneous measurements of temperature and relative humidity used for data correction. Consequent to the field testing of LCPMS at the stationary level, the sensors were integrated into an existing UA-hosted meteorological instrumentation suite. Air quality sensor performance was evaluated at Embry-Riddle Aeronautical University campus at Daytona Beach, FL by comparing data from the UA integrated LCPMS with the Vaisala AQT420 at the same altitude. Comparisons were made at different temperatures and relative humidity levels up to 400 feet above ground level at 50 ft increments. This novel method is tested by conducting flights in a sequence of increasingly complex scenarios designed to test interference of rotor-induced flow with the PM data. This work has implications for devising measures in urban, remote, and impractical locations.

Additional Authors: Andres Munevar, Marc Compere, Kevin Adkins, Marwa El-Sayed

Virtual Poster Discussion: Thursday, May 12, 8:00 a.m. PT

In-Person Poster Display: #3

Model Forecast Error Spatial Interpolation using Voronoi Tessellation in Ramboll Shair

Presented By: Ali Akherati, Ramboll

The boom of low-cost air quality sensors has driven a dramatic increase in air monitors. However, the air pollution concentrations can vary substantially, even between neighboring blocks. To understand air pollution trends on a local scale, Rambolldeveloped Shair, a real-time air quality monitoring system that models air pollution concentrations on a simple and accessible interface. The deployment in Richmond, California, democratizes access to reliable and hyperlocal air quality measurements and offers the community the opportunity to both understand air quality in real-time, and to identify major emission sources to support evidence-based emission reduction strategies. Shair incorporates a novel approach to use real-time spatial interpolation and sensor measurements to adjust the modeled concentrations on an hourly basis at a hyperlocal scale. Shair synthesizes publicly available sensor data, scientifically validated dispersion and weather models, street-level, regional physical and chemical transport models, and real-time traffic to estimate concentrations. The difference between measurements and modeled concentrations are compared at each point to calculate model forecast error at all observation points, and the residual model error is interpolated using Voronoi nearest neighbor inverse distance weightingto adjust the model’s estimate of air quality at every grid point, filling in the gaps in the network. The Voronoi polygon tesselations, recreated hourly based on available measurements, dynamically adapt to the real-time availability and density of the sensor network. Furthermore, Shair includes an automated review process using sensor operating specifications and historical trends, which flags suspicious measurements and possible outliers to prevent immediate integration with the model.

Additional Authors: Ayah Hassan, Justin Bandoro, Shari Libicki, Greg Yarwood

Virtual Poster Discussion: Thursday, May 12, 8:00 a.m. PT

In-Person Poster Display: #4

Towards a Global Low-Cost Sensor Calibration Model via Gaussian Mixture Regression

Presented By: Dan Westervelt, Columbia University

Low-cost Sensors (LCSs) for air quality monitoring have great potential to address the gap in air quality data across the globe. LCSs’ sensitivity to environment and source conditions mean they require the development of local calibration models built by collocation with reference grade monitors. Establishing sensor collocations and building local calibration models, however, is an expensive feat. Calibration methods like multiple linear regression (MLR) have been shown to effectively calibrate locally, but have failed to be applied across regions with differing climate and source conditions. Recent work in Accra, Ghana, a city with distinct climate regimes, has shown Gaussian Mixture regression (GMR) to be more successful than MLR when calibrating LCS data with heterogeneous characteristics (MLR R2= 0.81 & MAE = 2.8 µg m-3; GMR R2= 0.88 & MAE = 2.2 µg m-3).GMR is both a clustering and regression method that works by modeling the probability density of the output data conditional to the input data as a Gaussian Mixture Model (GMM).In addition to improving correlation and accuracy, GMR is shown to capture complex, physical relationships in data. GMM cluster assignments were shown to match underlying climate characteristics and seasonal trends, suggesting that a global GMM could have components for different pollutant mixes and climates and thus be applicable to multiple cities or regions. Building a GMM with data from just over 5 cities, and testing application on wide-spread global collocations via the CAMS-Net project, we have built a singular LCS correction model that can be effectively applied to any city globally. We validate the method by comparing our global correction factor to a local correction factor in several test cities. This global correction factor methodology could dramatically increase the utility of LCS data worldwide and could be incorporated into online open data repositories.

Additional Authors: Celeste McFarlane, Garima Raheja, Carl Malings, The CAMS-Net Team, Daniel Westervelt

In-Person Poster Display: # 7

Distant calibration for improved data quality from low-cost air quality sensors: a multi-testbed validation

Presented By: Jelle Hofman, Flemish Institute for Technological Research (VITO)

There is a demand for higher spatiotemporal and personalized air quality monitoring strategies to gain a better understanding on pollutant dynamics and improved representativity of people’s exposure to air pollution. IoT sensor technologies can meet those requirements but need proper calibration and interpretation in order to obtain reliable and actionable data. Our recent publication (Hofman et al., 2022) reports on a distant calibration procedure that is able to improve low-cost PM and NO2 sensor data using remote reference station data, i.e. without the need for recurrent co-location calibration campaigns. The algorithm accounts for sensor gain and offset, while compensating for observed sensitivities of low-cost optical (PM) and electrochemical (NO2) sensors. We report on different validation exercises using five different sensor testbeds (SDS011, OPC-N3, SPS30, NO2-A43F) at various urban and background locations in Belgium and the Netherlands. The results demonstrate that the distant calibration approach improves the sensor data quality (accuracy, linearity and correlation) up to sensitizing and supplementary (EU Class 1) data quality levels and can be scaled to different sensor networks. Thanks to its cloud implementation and openly available input data, this scalable calibration can be provided “as a service” on top of existing sensor networks. Although this approach has shown to improve sensor data, the ultimate performance will still depend on the applied sensor type, unit (design of sensor box) and granularity of the available reference monitoring network. Reference: Hofman, J., Nikolaou, M., Shantharam, S.P., Stroobants, C., Weijs, S., La Manna, V.P., 2022. Distant calibration of low-cost PM and NO2 sensors; evidence from multiple sensor testbeds. Atmos Pollut Res 13, 101246.

Additional Authors: Jelle Hofman, Mania Nikolaou, Sharada Prasad Shantharam, Christophe Stroobants, Sander Weijs, Valerio Panzica La Manna

In-Person Poster Display: #20

Air quality use cases: assessing the impact of different events using air quality data from model and sensors networks

Presented By: Jill Chevalier, eLichens

Road traffic is a main target source of action to improve air quality (AQ). The monitoring and evaluation of the impact of emission abatement actions on AQ is essential to better inform decision-makers. However, short-term, or localised actions remain very complex to evaluate without the implementation of significant means of modelling and measurements allowing to consider, the impact of meteorological variability on the concentration levels. High-resolution, continuous data on pollutants concentrations are needed to observe possible impacts of changes in mobility policies. However, model-only cannot reflect the reality accurately enough, as the local and short-term changes due to peculiar events might not be caught. A dense network of station is of paramount importance to catch such events. eLichens developed a mapping method to provide hourly, street-level data on urban areas. The model uses a neural network pollution model improved by fusing measurements from a network of eLos (eLichens outdoor station). eLos is a real-time recalibrated low-cost AQ station, providing hourly measurements of NO2, O3, PMs and CO2, temperature and humidity. Using this data, we assessed the impact of different events on AQ: Covid-19 lockdowns, summer fires or changes in local mobility policies. For long-lasting events, we compare the present data to the data from the previous year to get rid of the meteorological variability. We detected a drop in nitrogen dioxide concentrations in several cities during lockdowns. In Grenoble, France, the creation of a bike lane changed the traffic, creating jams. However, we evidenced a decrease in pollution levels.

Additional Authors:

In-Person Poster Display: #12

Impacts of air pollution on the Heath of inhabitants in the city of Douala: CAMEROON

Presented By: Robert MBIAKE, University of Douala

IMPACT OF AIR POLLUTION ON THE HEALTH OF INHABITANTS IN THE CITY OF DOUALA: CAMEROON The poor quality of the air in must Sub-Saharan cities becomes over time, a normal human’s life despite the worst impact on their health. They are two main causes that pollute the air. In one hand, the increasing of these cities at their breakneck speed and their rapid growth with an anarchic urbanization which is obviously accompanied with a number of second-hand cars (70%), motorbikes (300 000 at Douala city) with the dirty fuel constituted with crude oil where there is naturally lead, high sulphur. In the other hand, the air is also polluted there by the use of wood or charcoal for cooking in households and for the commercial food along the roadside. The main exposure persons are those permanently in contact with this air. For this work which will be the subject of our presentation, it consists in establishing the link between poor air quality and some diseases. First of all, we leaded previously some research to ensure that the city of Douala was really polluted. This research showed that unfortunately the concentration of PM10 could reach in some places 8 times higher than the WHO threshold while the PM2,5 concentration is 7 times (183,43 µg/m3) higher than the WHO threshold (25 µg/m3). Following this confirmation, we built two groups, one for outdoor workers constituted with motorbike riders, small traders congregated at the crossroad for their activities, fuel station service sellers and the second group was made up with indoor workers constituted by shopkeepers and storekeepers and women householders. We submitted a questionnaire formulated in accordance with British Medical Research Council (BMRC) and the American Thoracic Society Disease Lung Division ATS-DLD 78-C. Actually, 1,500 participants have responded to our questionnaire. Over the 16 clinical symptoms identified, 07 were regularly cited with the following frequencies:Colds (84.38 ± 1.6%); dry cough (75.86 ± 1.6%), headache (74.24 ± 1.7%), stinging eyes (66.29 ± 1.5%), general tiredness (63.07 ± 1.7%), runny nostrils (53.31 ± 1.1%) and watery eyes (52.38 ± 1.2%). The analysis of theses collected clinical manifestations was made from multivariable logistic model, Fischer and Pearson’s test, standard deviation and biostatistics analysis. The obtained results are suitable. It establishes the existence of a correlation between age and symptoms felt between smokers, alcoholics and clinical manifestations leading us to consider them as a cofounding factor. We have identified among these diseases those which could either be caused or accelerated following the breathing of polluted air. Keywords: Air Quality – Pollution – Clinical Manifestation – exposed persons – Biostatistics.

Additional Authors: MBIAKE Robert

In-Person Poster Display: #19

Determination of Particulate Matter in the City of Nairobi, Kenya Using Satellite Remote Sensing

Presented By: Moses Njeru, University of Nairobi

Particulate matter (PM2.5) has been known to have adverse health effects and to impact climate and our environment. In Nairobi, the capital city of Kenya, PM2.5 pollution has become a serious environmental problem due to the rapid industrialization, urbanization and expansion. However, because of scarcity in air monitoring stations, there is limited data and this inhibits policy making in the city. Satellite data might offer a viable solution to this problem by filling gaps between monitoring stations data. This research aims at using satellite data and data from low cost sensors to derive the surface concentrations of PM2.5 in Nairobi County. Precisely, satellite AOD obtained from MODIS sensor will be retrieved and associated to ground measurements of five optical particle counters in five areas in Nairobi County: Ruai area (mining and quarrying), Viwandani area (industrial site), University of Nairobi (urban area), Dandora area (waste disposal) and Kiamumbi (Background site). Land Use Regression (LUR) statistical model will then be used to analyze variability of the data correlated with land use data obtained from the Geographical Information System (GIS) data and meteorological data local weather station at the University of Nairobi.

Additional Authors: Michael Gatari, Moses Njeru

In-Person Poster Display: #14

Field calibration and evaluation of an IoT enabled low-cost particulate matter sensor in Birmingham UK

Presented By: Nicole Cowell, University of Birmingham

We evaluate the performance and calibration of an Internet of Things (IoT) enabled array of particulate matter sensors in Birmingham, UK. Commercial sensors (Plantower PMS5003) were adapted to utilise Low Power Wide Area Network IoT technology enabling battery power. The devices are capable of measuring & communicating data to an online platform with a battery life of ~2 months, at a measurement interval of 15minutes, allowing for automated air quality monitoring for extended periods at high density. In initial testing, sensors demonstrate success at being integrated into a wireless sensor network, with a high reliability of readings. The average correlation coefficients (r2) between raw PMS5003 data & reference instrumentation are 0.718, 0.703 and 0.543 for PM1, PM2.5 & PM10 respectively. The devices also demonstrate good inter-sensor consistency, with Pearson’s r values between pairs ranging from 0.92-0.99. As previously highlighted in the literature, relative humidity influences the response of the sensors, especially for RH > 85%. This was overcome by the development of a multi-linear correction factor for humidity effects. Using this model, Pearson’s r values range from 0.81-0.91 compared to 0.73-0.85 from uncorrected values. However, there are indications of drift at high humidity at ~8 weeks deployment time (in line with the battery life of the device). Finally, the Limit of Detection calculated from this study also demonstrates that the sensors are suitable for capturing concentrations typical of an moderately polluted UK urban environment- LoDs of PM2.5 in this study would have allowed for capture of 94.7% of the concentrations recorded at a typical UK urban roadside monitoring site between 2017-2020. Since this initial study, sensors have been deployed as part of a network across an urban environment. Preliminary data and sensor evaluations will also be presented from this initial network deployment. A live web platform for sensor data can also be viewed.

Additional Authors: Nicole Cowell, Lee Chapman, William Bloss, Francis Pope

In-Person Poster Display: #15

Evaluation of On-Campus Tree Planting Impact on Community Air Quality Using Low-Cost Sensor Measurements

Presented By: Ningxin Wang, Sonoma Technology

Legislative rules over the last five years (e.g., AB 617 by CARB and Rule 1180 by South Coast AQMD) have led to an increase in community-level air quality monitoring. Although agency-managed sites with regulatory air monitors are designed to protect public health by measuring ambient air within communities and ensuring that federal and state air-quality standards are met, they can be limited spatially when monitoring a specific community facility such as a school. Technological advancements in “low-cost” air quality sensors provide a more affordable and portable option for community air monitoring. In this study, we explore the impact of tree planting on community air pollution levels at two near-road schools in Fresno using low-cost sensor measurements. Fresno is among one of the most polluted areas in the U.S., frequently exceeding both the California and U.S. EPA ambient air quality standards for concentrations of fine particulate matter (PM2.5). Tehipite Middle School and Leavenworth Elementary School are both located in areas that are in the highest 10% of CalEviroScreen percentiles for census blocks in California and in close vicinity to major highways - California State Routes 99, 180, and 41. Several species of trees are planted along the fence between the school and adjacent major roads, and along the playground boundary. Previous studies have shown that vegetation can decrease ambient black carbon (BC) and/or PM2.5 downwind of the vegetation. We deploy low-cost sensors (i.e., Clarity node S), micro-aethalometers (MA350, Aethlabs), together with meteorological equipment at multiple sites on campus, and collect data on PM2.5, nitrogen dioxide (NO2) and BC concentrations. Approximately one month of data are collected before and after tree planting. Pollutant levels are evaluated together with meteorological conditions before and after tree planting. In addition, sensor performance is explored by comparing data among sites and between sensors.

Additional Authors: Ningxin Wang, Mona Cummings, Steve Brown

In-Person Poster Display: #16

Micromitigation: a Citizen Science Project for Volatile Organic Compound Adsorption in Ambient Air using Activated Carbon

Presented By: Rebecca E. Skinner, Counter Culture Labs

The session describes Micromitigation, which uses granulated activated carbon (GAC) to mitigate VOCs. The threat to human health posed by VOC toxic air contaminants is under-appreciated. The Micromitigation Working Group, hosted by Counter Culture Labs in Oakland, California, seeks to establish an open-source protocol to abate VOC air pollution. We place small screened panels of GAC in hotspots, then have the panels desorbed, contaminants incinerated, and the material reactivated. Conditions which facilitate adsorption in ambient air, rather than in closed canisters, have not been studied, and the mixing ratio of ambient air is idiosyncratic. The group is testing three strategies to adsorb effectively despite the low partial pressure of pollutants in ambient air. These are: repeated flows to adsorbent material, achieving saturation over a longer duration; deployment locations which maximize acute air pollutant concentrations; and increasing adsorption surface area with panels open to the air. In our initial experiment, screened panels of 4x8 mesh coconut-shell GAC placed in emissions hotspots in San Francisco and Oakland were tested at an analytical laboratory by TD-GC-MS analysis using a modified EPA Method TO-17 protocol. The results indicated significant hydrocarbons, especially aliphatic HCs as well as toluene, and long-chain HCs. Even such preliminary results indicate that passive ambient adsorption over a period of time is effective. The GAC was not blinded by PM 2.5, nor did competition from H20 preclude adsorption of VOCs. This process can be improved, and we seek new participants to work with us. This method could empower communities to mitigate a long-ignored carcinogen threat. Micromitigation could could be carried out by community groups with consultation from university or high-school chemistry teachers; and could be a useful demonstration project for environmental sciences, chemistry, and sensors and computing education

Additional Authors: Rebecca E. Skinner

In-Person Poster Display: #18

Utilising a co-ordinated and high-density low-cost sensor network for emission classification in an urban setting in Ireland

Presented By: Rósín Byrne, School of Chemistry, University College Cork

High density air quality sensor networks offer a unique opportunity to better understand the hyper-local and temporal patterns of air pollutants in an urban setting. With the co-ordination of reference instrumentation in the area, they are even more useful. This work builds on the established network of low-cost air quality sensors (LCS) measuring PM2.5 in an urban area in Ireland by combining the network with the official reference sites to create a powerful data set. The urban location contains four reference sites, however only two capture PM2.5 data. The data provided from these sites, while useful and necessary, does not provide information about hyper-local variations in pollution in the area. The first of its kind in the Republic of Ireland, this network of LCS was established by local authorities to better understand the local variations of particulate matter pollution in the area.

First, a calibration was applied to the sensors in order to bring them in line with the reference data. Linear regression and random forest calibrations were investigated and assessed with model validation techniques. Subsequently, the calibrated network was used to investigate the local variations and temporal trends in pollutant concentrations. By combining this robustly calibrated and co-ordinated network with local meteorological information, timeseries analysis and clustering techniques were employed to investigate the daily trends of pollution across the network and to identify and potential source regions, e.g. areas of accumulation and assessing if specific areas influence surrounding parts of the city.

Additional Authors: Rósín Byrne, Stig Hellebust

In-Person Poster Display: #11

Assessing PM species, NO2, O3, and Black Carbon Emissions from Agricultural Sugarcane Stubble Burning Episodes in La Feria, South Texas, USA

Presented By: Amit Raysoni, The University of Texas Rio Grande Valley

Emissions from agricultural stubble burning is a major source of air pollution in many developing countries of the world. However, in South Texas, USA, sugarcane stubble burning is a fairly common practice from October to May. Emissions from such activities on both sides of the U.S.-Mexico border can result in detrimental health effects for the neighborhood communities. The findings presented herewith are from an ongoing study conducted in the town of La Feria, South Texas - an area very close to the international border with Mexico. The instruments measuring various PM species, NO2, O3, and BC were deployed in a ranch-house near a sugarcane field. to characterize pollutant concentrations for about five weeks before and after the stubble burning activities. A low-cost TSI Blue Sky Sensor was also deployed to draw comparisons of measurements recorded by a low-cost sensor and other instruments mandated by the EPA as optimal for air pollution measurements. Findings from this study will be presented at the conference as this is an on-going study and the field sampling will end by the first week of April 2022. This study is very important for a low-resourced community such as La Feria due to the sheer paucity of adequate Central Ambient Monitoring Stations (CAMS). Texas Commission on Environmental Quality only operated five such stations in the whole of Rio Grande Valley comprising of four counties. There is one air monitoring station each in Brownsville, Harlingen, Edinburg, Mission, and Los Fresno near South Padre Islands. Out of these 5 CAMS sites, only two measure PM2.5 and one measure NO2. As such, such TCEQ CAMS sites are not an accurate representation of citizens' exposure burden and this very well accentuates the fact that on-site monitoring of pollutants in such remote communities is essential. This study, the authors believe, will definitely add to the body of air quality literature in this Hispanic majority region of South Texas.

Additional Authors: Edward Robles, Esmeralda Mendez, Amit Raysoni

In-Person Poster Display: #86

Community Air Sensor Use

A citizen-science approach to monitor air quality in a UK school adjacent to trafficked roads during lockdown

Presented By: Hamid Omidvarborna, University of Surrey

Many of the UK schools are located adjacent to busy roads for better accessibility. Infiltration of the emissions from road vehicles disproportionately endangers the health and wellbeing of vulnerable young children. The COVID-19 lockdown restrictions resulted in limited mobility in the UK, however, step-by-step ease in lockdown led to a change in the exposure level of the school children. This citizen science study, which is a collaboration between parents, nearby residents, school personnel (collectively as citizens) and the Guildford Living Lab (GLL) at the University of Surrey, was initiated before school reopening when ‘the temporary reduction in urban air pollution during the COVID-19 forced confinement’ topic was of great interest. The objective is to assess the impacts of the UK lockdown phases on ambient air quality in a primary school. The ambient air quality monitoring station was installed inside the school in December 2020 and the concentrations of particulate matter in different size fractions (PM1, PM2.5, and PM10) and gaseous pollutants (e.g. NO2, O3and CO) were monitored. Polar plots showed that the concentrations varied across each phase of the lockdown and started to increase when restrictions were lifted. We explored the relative importance and levels of the explanatory variables for each pollutant and later removed the impact of weather to deeply assess the extent to which changes in ambient pollutant concentrations were attributed to changes in emission levels following the lockdown eases. Our results revealed the role of domestic transport use (e.g. heavy goods vehicles and light commercial vehicles) in detrotiating ambient air quality, therefore, undertaking additional analysis, like source apportionment, is suggested.

Additional Authors: Hamid Omidvarborna, Prashant Kumar

In-Person Poster Display: #78

Using low-cost sensors for hyperlocal PM2.5 assessment: lessons learnt from two experiments.

Presented By: Jalal Awan, Pardee RAND Graduate School

According to the WHO, air pollution is the largest single environmental health risk, estimated to kill 1 in 8 people globally, due to heart disease, stroke, respiratory disease and cancer. Moreover, blacks and Hispanics are often more exposed to air pollution than whites, despite contributing less towards producing it. Community-based air quality monitoring provides an interesting avenue for research. The widespread diffusion of Internet-of-Things (IoT) connected devices, high-speed telecommunication networks and open data are reconfiguring the way data underpinning policy and science are being produced and consumed. One such recent paradigm in air quality monitoring is the use of low-cost sensor networks for more granular, neighborhood-level air quality assessments. In my poster presentation, a summary of current paradigms in air pollution monitoring are presented along with a discussion on key findings from two experiments deploying 30 low-cost air quality sensors in the City of Santa Monica and Pittsburgh. RAND researchers based in the two cities were chosen as a convenience sample and correlations of PM2.5 across various dimensions (environmental variables, time of the year / day / month, wildfire events etc) over a 2 year deployment period were compared. Based on findings from our experiment, I provide policy level guidance for federal and local agencies aimed at expanding research and use of community-science based air quality monitoring. Further discussion (and my dissertation topic) based on our findings on performance evaluation of low cost sensors in field and calibration methods against EPA’s FRM / FEM is also presented.

Additional Authors: Jalal Awan, Christine Chen, Sangita Baxi, Alejandro Becerra

In-Person Poster Display: #74

A Low-Cost Industrial Particulate Profiler Simultaneously Reporting PM10 and PM2.5

Presented By: Jennifer Brown, Met One Instruments, Inc.

Met One Instruments, Inc. has developed an industrial-grade optical particulate profiler that simultaneously measures and reports ambient particulate matter at PM10, and PM2.5cut points. This device, known as the “ES-412 Simultaneous Particulate Profiler,” can operate for up to two months without user intervention and will report these parameters with hourly sensitivity of less than 0.1 mg/m3. The ES-412 offers wireless, remote PM monitoring in a low-profile, field-deployable, weatherproof unit. It is lightweight and self-contained. The ES-412 provides several features and diagnostics, such as active flow control, generally not found in low-cost PM sensors. The ES-412 is a complete system with integrated cellular communications, a customized webpage dashboard, a 3-year data plan, an AC power supply, and a transport case. Users have instant access to data on their smartphone, tablet, or computer. Data from the ES-412 is backed up internally and then transmitted to a nearby cell tower every 15 minutes. In the event of cell tower communication loss, the ES-412 will attempt communication continuously until it is regained. When this occurs, the ES-412 will transmit all missing data when communication is re-established, providing data security. This presentation will report field test results from several field sites operating collocated, US-EPA-designated PM10 and PM2.5monitors (Met One Instruments BAM-1020). We will provide accuracy and relative standard deviation (CV) measurements that demonstrate near equivalency for factory-calibrated ES-412 monitors. Local span calibrations are not allowed on an EPA-designated PM monitor. However, if the ES-412 is span calibrated by collocating it with an EPA-designated monitor or sampler, it can be demonstrated that accuracy can generally be maintained for significant periods of time. We will also report how the ES-412 has benefitted local school districts in the area during wildfire season.

Additional Authors: Jennifer Brown, David A. Gobeli, Ph.D., Leah Miller

In-Person Poster Display: #76

Sensors, Flags, and Libraries on the Nez Perce Reservation

Presented By: Mary Fauci, Nez Perce Tribe Air Quality Program

The Nez Perce Tribe (NPT) Air Quality Program has a robust partnership with nine public libraries on the Nez Perce Reservation. Our partnership started in 2012 through participating in the libraries’ annual Summer Reading Program. In 2021, we expanded our collaboration to promote Smoke Ready Communities and each library received a PurpleAir sensor and an AQI flag kit. In addition to hosting a PM2.5 sensors for the tech-savvy, library staff took on flying the AQI flags outside. Libraries purchased supplies, built, and displayed DIY Box Fan Filters to improve their buildings’ indoor air and demonstrate a proactive step residents could take to protect themselves from wildfire smoke. Most recently, two libraries are piloting a Moisture Meter Loan Program for firewood users. We will also share our evaluation of sensor performance comparing library data with our NPT Ambient Network PM2.5 monitors (BAM 1022s and EBAMs). Our librarians have become AQ ambassadors. It’s a win- win partnership.

Additional Authors: Mary Fauci, Johna Boulafentis

In-Person Poster Display: #77

Evaluating personal exposure in a Vancouver neighbourhood through community collaboration and the implementation of a low-cost sensor network

Presented By: Rivkah Gardner-Frolick, University of British Columbia

Exposure to air pollution is dependent on a person’s daily movement through microenvironments. These microenvironments reflect the strong contribution of local sources to community air quality patterns. To quantify small-scale spatial gradients in communities concerned about air quality, monitoring campaigns often use low-cost sensors (LCS). One such community, the Strathcona neighborhood in Vancouver (Canada), contains a variety of industrial and transportation emissions sources within its boundaries. In addition, Strathcona is home to many sociodemographic groups that are especially vulnerable to air pollution, such as Indigenous residents, unhoused residents, and the elderly. Implementing a network of LCS in this area to map fine-scale air quality in combination with surveys of time-activity patterns will help to assess air quality patterns, personal exposure, and potential exposure mitigation strategies. This study is a collaboration between academic researchers and the Strathcona community. The Strathcona Residents Association is the main community representative and has contributed to sensor network design, spatial mapping, data interpretation, and data visualization. We are assessing air quality by deploying 16 LCS units (SENSIT RAMP) across Strathcona for a period of 6 months to measure PM2.5, CO, and NO2. Each monitor is placed where there is community concern about high-emitting sources or vulnerable populations. Spatial mapping for each season will use land use regression and incorporate community knowledge of sources and receptors, enabling better estimation of small-scale sources and quantification of the value added by community input. Visualization of the spatial map will be done collaboratively and will include representation of sources of concern and vulnerable receptors from Strathcona residents in addition to air pollution concentrations. The collaborative map will produce easily accessible information for residents and practitioners.

Additional Authors: Rivkah Gardner-Frolick, Sakshi Jain, Nika Martinussen, Dan Jackson, Trefor Smith, Emily Peterson, Naomi Zimmerman, Amanda Giang

In-Person Poster Display: #75

Using Low-Cost Sensors to Assess Intra-Urban PM Species Concentrations at Lower Rio Grande Valley Region of South Texas

Presented By: Amit Raysoni, School of Earth, Environmental, and Marine Sciences at The University of Texas Rio Grande Valley

Given the increasing developments in ambient monitoring using low-cost air quality sensors, twelve BlueSky Air Quality Monitors (TSI Incorporated, Minnesota, U.S) were deployed in various towns across the Lower Rio Grande Valley (RGV) Region of South Texas from March 01, 2021, to March 31, 2022. Spatial and temporal variability in particulate matter species (PM1, PM2.5, PM4, and PM10) concentrations were analyzed using real-time hourly and 24-hour data concentrations from these sensors. The assessment of PM from the BlueSky monitors was compared to the PM2.5 data at the three RGV Texas Commission on Environmental Quality (TCEQ) Continuous Ambient Monitoring Stations (CAMS) sites (C43, C80, and C323). Central CAMS sites are limited in the RGV area; therefore, the neighborhood-level ambient monitoring with the low-cost sensors has the potential to accurately represent actual exposure patterns in the community. Coefficients of Divergence (COD) and Spearman’s Rho correlational analysis is conducted for spatial and temporal variability between sensors data and CAMS monitoring sites in the region. Hot-Spot Analysis (Getis-Ord Gi) and Cluster and Outlier Analysis (Anselin Local Moran’s Index are computed to assess PM exposure patterns. This study is important because it is one of the first to assess and characterize PM in many cities and towns in the lower RGV region using low-cost sensors. The results of this research will aid policymakers in creating guidelines for citizens to limit their PM exposures.

Additional Authors: Esmeralda Mendez, Dawid Wladyka, Katarzyna Sepielak, Owen Temby, Amit Raysoni

In-Person Poster Display: #85

Filling in the air quality data gap and enabling air quality management in LMICs using low-cost sensors

Experience of Mobile Air Quality Monitoring with IoT technology in the Historic Center of Lima.

Presented By: George Castelar Ulfe, Municipalidad de Lima

One of the biggest issues facing the city of Lima is the poor air quality. According to WHO (2005), the air exceeds by 2.4 times the recommended values for PM2.5/PM10. In this sense, Lima's historic centre is no exception to this problem. Therefore, it is necessary to develop more research studies in order to determine the main sources of pollution. The objectives of this study are to evaluate the temporal and spatial pollutant concentration, identify the areas with the highest concentrations of pollutants, and determine the amount of exposed vulnerable population. The air quality measurement used IOT technology installed in an electric vehicle, and was performed around the historical area called “Damero de Pizarro”. The study involved measurements of meteorological parameters (relative humidity and temperature), and the concentrations of 7 atmospheric pollutants (PM2.5, PM10, CO, O3, NO2, SO2and H2S) during the months of June, July and August 2021. The results were considered in 4 shifts: 8:00 to 10:00; 10:00 to 12:00; 14:00 to 16:00 and 16:00 to 18:00 hours. The obtained baseline values did not show an excess of the limits established in the national environmental air quality standards. However, the average concentrations of PM2.5and PM10were relatively high during the first 2 shifts (8:00 am to 12:00 pm). In addition, 18% of the exposed population was found to be the most vulnerable to air pollution. Moreover, it was also found that the highest concentrations of pollutants were found on the roads with the most traffic. Finally, this study generated empirical evidence that allows the local government to promote actions in favour of improving air quality of its citizens. Projects like the pedestrianisation of streets in the study area and the implementation of a future Low Emission Zone (LEZ) are promoted within the framework of the "Action Plan for the Improvement of Air Quality in Lima and Callao 2021-2025".

Additional Authors: George Castelar Ulfe, Erick Ingaroca, Cristina Rentería

In-Person Poster Display: #95

A Low-Cost air quality system for long-term monitoring: the BOCS

Presented By: Sebastian Diez, University of York

Every year ~9 million people globally die prematurely because of air pollution, with the poor and vulnerable disproportionately affected: 92% of pollution-related deaths occur in low and middle-income countries (LMICs). Furthermore, regardless of the country's income, air pollution health effects are overwhelmingly prevalent on minorities and the marginalized. But due to the lack of resources in many LMICs long-term monitoring using traditional reference-grade instrumentation is prohibitive, and the concentrations, sources and effects of air pollution are poorly understood. Low-cost sensors (LCS) could potentially fill this gap by providing the necessary air quality data to inform local mitigation. Unfortunately, the majority of LCS systems currently available are “black-box” in nature, making them difficult to maintain in-country and expensive to operate long-term. In this study, we present a new fault-tolerant and easy to maintain instrument, the Box Of Clustered Sensors (BOCS), which combines clusters of LCS to redundantly quantify CO, NO, NO2, Ox, CO2 and VOCs in a modular design, in order to provide robust and reliable measurements of air pollutants. The open-source design and calibration algorithms used in the BOCS significantly reduce operation and maintenance costs, enabling pollutant monitoring in LMICS to be performed by local air quality managers. This approach also provides a transparent and reproducible data processing pipeline, in order to deliver reliable uncertainty estimates on the data provided. In this presentation we will introduce the BOCS and the simple calibration models applied for CO, NO, NO2 and O3 employing the multiple BOCS signals.

Additional Authors: Sebastian Diez, Stuart Lacy, Killian Murphy, Pete Edwards

In-Person Poster Display: #96

Performance evaluation of low-cost electrochemical sensors for ozone in two polluted urban areas.

Presented By: Pamela Ayala, Airflux SPA

Low-cost electrochemical sensors (LCS) were used for ozone monitoring in two highly polluted urban areas, one corresponding to the capital and main city of Chile, and the other to an industrial coastal zone. This secondary pollutant is formed from photochemical reactions of primary pollutants present in the atmosphere, through non-linear processes. Due to their complex formation dynamics, it is important to evaluate the performance of Alphasense LCS, which combine a nitrogen dioxide (NO2-A43F) and oxidant gas (OX-A431) sensor as a pair, in two zones that present weather conditions. , topographic and different emission sources. The voltage outputs registered by the sensors were transformed to concentration units, through a multiple linear regression model, where the predictors considered as temperature, reluctant humidity and interfering pollutants, depended on the area where they were installed. Through the data in concentration units, the performance of the sensors is evaluated according to objective values of statistical metrics such as mean square error, standard deviation, among others; which are stipulated by the US Environmental Protection Agency; it is also expected to characterize the dynamics of ozone in different areas and its precursors.

Additional Authors: Sebastian Diez, Pete Edwards, Stuart Lacy, Killian Murphy

In-Person Poster Display: 111 - Virtual Only

Indoor Sensing for Air Quality Control and Ventilation Applications

Development of a bespoke sensing unit for deploying in smart homes

Presented By: Hamid Omidvarborna, University of Surrey

The substantial developments of air quality sensors for indoor monitoring offer a great potential to ensure healthy living for residents. The use of air quality sensors is also included in the concept of smart homes and further connected homes to develop smart neighbourhoods or cities, where innovative technologies are implemented to proactively address a number of factors including the health and wellness of inhabitants. Building trust in the quality of data for making informed decisions at different scales requires proper data collection. Therefore, elements of air quality sensors should be properly selected and their performance should be rigorously evaluated. Here, we selected, designed and manufactured a bespoke unit to measure various indoor elements, including but not limited to, temperature, relative humidity (RH), noise, particulate matter (PM) in different mass fractions (PM1, PM2.5, and PM10) and size distributions, and carbon dioxide (CO2) for the aim of deployment in MyGlobalHome prototype building at the Innovation Centre and later at the pilot demonstration building, University of Surrey, Guildford, the UK during pre- and post-occupancy. The sensors are evaluated inside the state-of-the-art Environmental-Pollution (Envilution®) Chamber under the conditions that each sensor might experience inside homes. The methods for performance evaluation are elaborated in Omidvarborna et al. (2020). The correlation coefficients during performance evaluation are promising, where strong correlations for temperature (R2=0.99), RH (R2=0.99), PM1(R2=0.93), PM2.5(R2=0.92), PM10(R2=0.91), and CO2(R2=0.99) were obtained under-represented indoor environments. For further evaluation, the sensors are currently being integrated into a single unit with other communication features to become a part of the prototype building during the trial test. After successful evaluation during a long term occupancy trial, the sensors will be integrated into the design of smart homes.

Additional Authors: Hamid Omidvarborna, Prashant Kumar

In-Person Poster Display: #79

Understanding heterogeneity and sources of black carbon in residential neighborhoods in the Capital Region of New York State

Presented By: Md Aynul Bari, University at Albany

With recent findings of improving outdoor air quality due to current COVID-19 pandemic, there is an interest in understanding the potential impact on indoor air quality. Black carbon (BC) is a potent short-lived climate pollutant and an important component of particulate matter emitted from fossil fuel combustion and biomass burning (e.g., wood stoves) and has linked to adverse health outcomes. In the United States, current observation networks for BC are limited in characterizing exposure across neighborhood scales. Little information is available about BC concentrations at indoor environments. To address this gap, an exploratory study has been conducted to determine spatiotemporal variation of indoor and outdoor concentrations of BC, identify indoor-generated and potential local source impacts in New York State Capital Region covering urban and rural residential neighborhoods including Environmental Justice (EJ) communities. Indoor and outdoor sampling has been performed in 4 homes of each selected neighborhoods to collect data from at least 20 homes. We leveraged both low-cost sensors and microAeth instruments to measure BC. Indoor measurements are taken at a breathing height (~1.5 m) within the living room, while outdoor sampling are deployed in backyards. Data on meteorological parameters e.g., outdoor wind speed, wind direction and relative humidity, temperature, as well as questionnaire-based housing characteristics, and occupants’ activities were also collected for each home. Our preliminary data suggests that outdoor concentrations (at backyards) were significantly higher than indoors suggesting an influence of potential local sources. Significant variations in indoor concentrations were observed among neighborhoods depending on the location of homes. The findings can benefit the general people to improve their knowledge, raise awareness and empower communities to take actions and inform policy makers to improve indoor air quality and public health.

Additional Authors: Sanchita Paul, Marco Eugene, Rafael Pereira, Zahirul Khan, Md Aynul Bari, Brian Frank

In-Person Poster Display: #80

Assessment of PM2.5 concentration and transport in indoor environments using low-cost air quality monitors

Presented By: Sumit Sankhyan, University of Colorado Boulder

Fine particulate matter (PM2.5) is an important constituent of air pollution and has been linked to a variety of health effects.Consumer-grade, low-cost PM sensors are gaining popularity as a convenient tool for consumers to monitor indoor air quality inside their homes. We investigated five commercially available air quality monitors (IQAir AirVisual Pro, Foobot Home, PurpleAir PA-II-SD, and PurpleAir PA-I-Indoor) and compared them to a research-grade optical particle monitor (OPS 3330, TSI Inc.) by deploying them in four homes of different sizes over a period of 9-12 weeks each. Two identical units of each monitor were deployed in the kitchen and bedroom of each home to evaluate PM2.5 transport between those spaces. Indoor monitors were collocated for 3 days at the beginning and end of each deployment period to assess their accuracy over time.A second component of the project included an investigation on the effects of deploying a consumer-grade portable air cleaner in the kitchen and in the bedroom on PM2.5levels.Preliminary results show a range of correlation levels between low-cost monitors and the OPS, with R2 values ranging from 0.75 to 0.94, and among pairs of the same low-cost monitor models, with R2 values between 0.60 and 0.99.

Additional Authors: Sumit Sankhyan, Julia Witteman, Steven Coyan, Sameer Patel, Marina Vance

In-Person Poster Display: #82

Cloud connected sensor netowrks and building reopening in 2022

Presented By: Timothy Quinn, SGS Galson

Covid-19 has disrupted indoor work environments and occupancy levels worldwide. We will examine the science behind airborne transmissions of virus’ new mitigating strategies. We shall examine 4 things common in indoor office spaces, by monitoring their levels with new cloud-based sensor technology we can help predict the likelihood of their presence in these spaces and possible transmission to workers. What responsibility does industry have to provide safe work environments, clean and disinfect buildings and use real time monitoring equipment? When a mesh sensor network is deployed building managers and system operators can gain real-time values and alerts that can keep them informed about the operation of interventions implemented to mitigate the transmission of COVID-19 throughout the workspace and keep the most vulnerable spaces in the building operating safely. Real-time monitoring creates a framework to provide the assurance of a healthy indoor environment and the confidence people need to resume their daily lives.

Additional Authors: Timothy Quinn

In-Person Poster Display: #83

Continuous Monitoring of Airborne Particles and Carbon Dioxide in Meat Processing Plants

Presented By: Mehael Fennelly, School of Chemistry and Environmental Research Institute, University College Cork

During the COVID-19 pandemic, meat processing plants (MPPs) have been vulnerable to transmission of SARS-CoV-2 with outbreaks documented in multiple jurisdictions. Several factors make viral transmission difficult to control in these settings including environmental conditions, the nature of the work, and difficulties in implementing physical distancing.

Boning halls were pre-selected as air monitoring sites based on documented evidence of the area having high proportions of the overall number of PCR-positive cases during outbreaks.

Air quality measurements were conducted in two MPPs with different air handling systems over a period of six months. Carbon dioxide (CO2) and PM2.5 concentrations, temperature and relative humidity were continuously recorded using an AirVisual Pro air quality monitor (IQAir, Staad, Switzerland). The number and size distribution of aerosol particles over the size range 0.75 – 12 μm was measured in real-time using a Wideband Integrated Bioaerosol Sensor (WIBS-4A; Droplet Measurement Technologies, Colorado, USA), which also uses fluorescence to identify the fraction of aerosol particles that are of probable biological origin.

The number of fluorescent particles and CO2 concentrations both varied according to the room occupancy and worker activity. Significant correlations were obtained between fluorescent particle numbers and CO2 concentrations at site A (R=0.40) and site B (R=0.81), suggesting that the fluorescent particles measured are respired biological particles

Ventilation rates were calculated from the exponential decay of CO2 were ~0.7 air changes per hour (ACH) in a non-ventilated boning hall and 2 ACH in a highly ventilated boning hall.

These results show the WIBS-4A and low--cost sensors can monitor and characterise a sensitive indoor working environment, detect change over time and potentially evaluate interventions empirically used to improve air quality.

Additional Authors: Mehael Fennelly, Donal Sammin, David O'Connor, Michael Prentice, John Wenger, Stig Hellebust

In-Person Poster Display: #81

Using Low-Cost Sensors to Quantify the Effectiveness of Portable HEPA Filter Air Cleaners for Reducing PM Exposure in Homeless Shelters

Presented By: Ching-Hsuan Huang, Department of Environmental and Occupational Health Sciences, University of Washington

Over three thousand portable air cleaners (PACs) with high-efficiency particulate air (HEPA) filters were distributed by Public Health - Seattle King & County to homeless shelters during the COVID-19 pandemic to control SARS-CoV-2 transmission. Although considerable evidence supports improvements in indoor air quality with PACs in residential settings, few studies of HEPA-based PACs in larger congregate settings such as shelters have been conducted. For these larger settings, where many air cleaners may be used, there is a need to quantify the usage of HEPA-based PACs and to correlate performance with site and building characteristics, and management decisions. In this study, we are interested in understanding the effectiveness of HEPA-based PACs in reducing airborne particulate exposures during the remainder of the pandemic across different homeless shelters. Multiple HEPA-based PACs were deployed at three homeless shelters in King County, Washington, for use in sleeping quarters. To quantify the indoor particle air quality improvements over the deployment, we conducted monitoring using low-cost optical particle sensors in each of the indoor zones of the shelters. One outdoor ambient location was selected at each shelter. Sensor data was collected for two-week deployments, separated by single-week gaps, over a three-month period. Continuous measurements of CO2 were also made in the indoor zones of the shelters with low-cost monitors to assess air exchange rates. The energy consumption of each PAC was measured using an energy data logger to allow tracking the PAC use and fan speed. The results indicate low indoor to outdoor particle ratios, despite varying PAC use in the three shelters. The CO2 levels in the sleeping dorms of the three shelters reveal diurnal patterns related to occupancy and ventilation and were generally within the specified American Society of Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE) guidelines for comfort (< 1000 ppm).

Additional Authors: Ching-Hsuan Huang, Jeffry Shirai, Elena Austin, Martin Cohen, Timothy Gould, Timothy Larson, Thu Bui, Daniel Huang, Shirlee Tan, Edmund Seto

In-Person Poster Display: #108

Low-Cost Air Sensors to Investigate Woodsmoke Interventions in an Exposure Chamber

Presented By: Lilian Liu, University of Washington

Wildfire smoke has emerged as a priority environmental health concern in recent years. Populations affected by wildfires suffer both economic loss and negative health impacts from smoke exposure. Infiltration of smoke deteriorates indoor air quality and increases risks of exposure to PM2.5 and subsequent health outcomes. Therefore, characterizing indoor exposures to wildfire smoke and providing practical and effective mitigation strategies are urgent needs.

Our study used low-cost sensors, a modified Dylos DC1700 and Met One AEROCET 531, to establish a woodsmoke particle concentration in an experimental exposure chamber and to characterize the effectiveness of using HEPA air purifiers for particle removal. To simulate wildfire smoke episodes, we employed a 333 ft3 tent as the exposure chamber and introduced smoke from cooking smoke guns and a wood pellet stove, respectively. Additional research instruments, including a TSI 3330 Optical Particle Sizer and TSI 3910 Nanoparticle Sizer were used to measure PM and ultrafine particles for a comprehensive particle size profile and for comparisons to low-cost sensors. The system was used to assess the effectiveness and efficiency of a HEPA air purifier at different fan speed settings for indoor particle removal.

Both low-cost and research-grade instruments demonstrated the feasibility of using a smoke-generation system paired with a tent for wildfire smoke exposure simulation. At high fan speed, the HEPA air cleaner achieved particle decay rates of 46.7/h for the newly installed filter and 38/h for the purifier with a used filter, supporting the use and regular replacement of HEPA filters as an effective approach to improving indoor air quality. Results comparing the air cleaner with new vs. used HEPA filters showed that the HEPA air purifier with new filters installed removed particles more efficiently.

Additional Authors: Lilian Liu, Chaja Levy, Alia Alhunaidi, Elena Austin, Shirley Huang, Jeffry Shirai, Edmund Seto

In-Person Poster Display: #109

Innovative Sensor Technologies

Performance Evaluation of the Auxiliary Electrode in Improving Data Quality from 4-pin Electrochemical Gas Sensors

Presented By: Anna Farquhar, Aeroqual Ltd

Electrochemical gas sensors (GSEs) are a low-cost option for the measurement of pollutant gases in outdoor environments. They are commonly deployed in low-cost sensor networks. The reliability of GSEs is impacted by changing environmental conditions, including temperature and humidity. Various strategies have been deployed to mitigate the impact of temperature or humidity changes, including the development of the auxiliary electrode (AE) in 4-pin sensors. The AE sits below the working electrode (WE) and responds to changes in temperature and humidity only. In this work we evaluate the performance of the AE in improving the output of carbon monoxide and nitrogen dioxide sensors from two manufacturers. An empirical algorithm that subtracts the output of the AE from the WE was used to correct for the steady-state temperature offset of the baseline concentration. However, the WE of a sensor shows large amplitude fluctuations (± 20 ppb) in response to rapid changes in humidity. In this work we show that the AE also responds to rapid changes in humidity, however the direction and magnitude of the AE fluctuations are different from the direction and magnitude of the WE fluctuations. By subtracting the AE output from the WE output the already large baseline fluctuations are magnified rather than mitigated in many cases. This is especially concerning for the NO2 sensor. NO2 concentrations in ambient environments are of a similar magnitude to the baseline humidity fluctuations, so cannot be discriminated from baseline fluctuations. In this presentation we demonstrate the limited usefulness of the AE in real-world applications and how if improperly implemented the AE can negatively impact data quality. We also describe alternative options for mitigating the effects of humidity that could be employed in low-cost sensor networks. *Selected for a Lighting Talk

Additional Authors: Anna Farquhar, Geoff Henshaw

In-Person Poster Display: #97

AROMA-ETO: Part-Per-Trillion Sensitive, Realtime Ethylene Oxide Measurements in Ambient Air

Presented By: Anthony Miller, Entanglement Technologies

Ethylene oxide (EtO) is a common chemical used in chemical manufacturing and commercial sterilization processes. It has recently received a large degree of focus for federal, state, and municipal governments as well as industry and community groups due to its increased potential cancer risk from long-term inhalation exposure. Lab quality, part-per-trillion measurements of EtO in air is needed to accurately identify areas of concern, understand background concentrations, and keep industrial workers safe. In this work, the development and performance validation of Entanglement Technologies’ AROMA-ETO will be described. AROMA-ETO is a thermal desorption cavity ringdown spectroscopy (TD-CRDS) analyzer, which delivers real-time, in-field measurements of EtO with detection limits in the low part-per-trillion range. This capability enables rapid field surveys and assessments that are required by policymakers and industrial operators to make time-sensitive decisions to reduce EtO-related health risks, including stopping leaks and releases before they can cause harm or non-compliance. The system is also ideal for measuring EtO in ambient air as it is able to accurately measure background concentrations and identify hot spots. Ambient air data of EtO concentrations collected using the AROMA-ETO will be presented alongside other use cases. Entanglement Technologies’ AROMA platforms are used for mobile monitoring, industrial EHS, fenceline monitoring, emergency response, tracking fugitive emissions, among others.

Additional Authors: Anthony Miller, Jake Margolis, Aurelie Marcotte, Michael Armen

In-Person Poster Display: #98

An array of low-cost PM sensors to characterize the structure of roadside microscale atmospheric flows

Presented By: Aron Jazcilevich, Universidad Nacional Autónoma de México

Street concentration fields change in time scales measured in seconds. This is because they are subject to immediate emissions and turbulence produced by vehicular wake. PM concentrations may vary from 20 micrograms/m3 to a spike of more than 1000 micrograms/m3in about 3 seconds. Furthermore, the distribution of pollutants on the roadside is uneven, generating local maxima concentrations spots depending on traffic, street design and existing meteorological conditions. These factors complicate the analysis of acute exposure events. An array consisting of 224 low-cost PM2.5sensors placed on the roadside obtains instantaneous concentration maps on a plane perpendicular to the road. This allows us to obtain the position and preferred height and distance to the road of maximum concentrations. The knowledge gained by using this array benefits the study of acute roadside exposure phenomena, location of specific risk zones, urban design, and in the evaluation of ecological barriers.

Additional Authors: Aron Jazcilevich, Luis-Darío Reyes, María del Pilar Corona Lira, Alberto Caballero, Antonio Suarez Bonilla

In-Person Poster Display: #100

Source apportionment of speciated VOCs with low-cost metal oxide sensors

Presented By: Caroline Frischmon, University of Colorado Boulder

Source apportionment of speciated VOCs allows communities to address pollution concerns in the order of highest impact to human and environmental health. Although regulatory-grade instruments are capable of VOC source attribution, their high cost prevents measurements on the spatial variability of these VOC sources. Performing source attribution with low-cost (LC) sensors instead opens the door to explore the spatial variability of local VOC hotspots because we can deploy the sensors as a network. This study deployed an array of LC metal oxide sensors into four Colorado communities that range from rural to urban, and which are all situated near oil and gas development. LC sensors were deployed for about one month, and these data were fit to regulatory-grade measurements of speciated VOCs using artificial neural networks (ANNs). Positive Matrix Factorization (PMF) grouped VOCs by their likely sources, such as wet and dry components of oil and gas operations, to begin discerning VOC source contributions and compositions at higher spatial resolution. Fits using ANNs had R2values of 0.6 for calibration despite concentration ranges that were orders of magnitude below the LC sensors’ prescribed detection limits. The fitted data captured baseline trends well but failed to estimate peaks. Grouping species by source using PMF generally improved fits, which we hypothesize is because the grouped concentrations were higher than that of individual species. PMF analysis revealed an oil and gas factor in each community; however, other source factors, such as biogenic and combustion, varied based on which reference species data were available in each site and what other sources are likely present in a rural, urban, or suburban environment. These results highlight the potential for VOC source attribution with LC metal-oxide sensors. Future works should seek to improve spike-finding and address challenges with species that are measured well below sensor detection limits.*Selected for a Lighting Talk

Additional Authors: Caroline Frischmon, Kristen Okorn, Michael Hannigan

In-Person Poster Display: #101

IEEE P2520™ – Standard for Testing Machine Olfaction Devices and Systems

Presented By: Ehsan Danesh, Advanced Sensing Technologies Ltd

A new series of international odour monitoring and analysis standards are being developed by the IEEE Sensors Council and the IEEE Industrial Electronics Society, in collaboration with the International Society for Olfaction and Chemical Sensing (ISOCS). The standards are designed for those developing and using odour analysis devices, electronic noses (e-nose), and Volatile Organic Compounds (VOC) analyzers. It is structured as a series of standards that target different, common, odour emission applications, such as outdoor air pollution and chemical processes. We believe that by targeting specific applications we hope to get more traction than previous attempts at standards. This paper gives a brief introduction to the series and solicit participation by individual scientists and engineers who see the importance of this work, by companies producing monitoring equipment, and by entities that set regulations that govern this field. Keywords: odour standards,artificial olfaction, machine olfaction,electronic nose (e-nose), VOCs detection

Additional Authors: Ehsan Danesh, Susana Palma, James Covington, Susan Schiffman, H. Troy Nagle

In-Person Poster Display: #102

UAV Deployed Sensors for Air Quality Monitoring

Presented By: Frederick Clowney, InterMet Systems

The availability of less expensive sensor options for air quality monitoring has created a new model for distributed sensor networks to supplement central monitoring stations. The InterMet X-4 uses the same approach for UAV deployed sensors, creating new opportunities for research and air quality management. Using low-cost UAVs, observers can take vertical measurements of the Atmospheric Boundary Layer to the extent of current restrictions (currently 400 feet). In additional, UAV deployment will allow for horizontal profiles that will increase the spatial resolution of fence-line monitoring, fugitive emissions and the measurement of local impacts of ship and vehicle emissions. InterMet Systems has been building high performance, low-cost atmospheric sensors for UAV deployment since 2011. The iMet-XQ, XQ2 and XF products have set the industry standard for small, lightweight Pressure, Temperature and Humidity (PTU) sensors. In 2020, InterMet began work on a new, UAV deployed sensor line that will measure pollutants as well as PTU. The modular design features a central processing hub that will be offered along with a family of integrated sensor modules. The X4 Hub will manage power and data and include four ports for redundant humidity and air temperature sensors. There will be two additional ports for gas sensor modules, the first of which will measure CO and CO2. Future modules are planned for O3/NO2, particulates and VOCs. This presentation will describe the physical and performance attributes of the iMet-X4.

Additional Authors: Nathan Dunn, Frederick Clowney

In-Person Poster Display: #103

Using the PurpleAir monitor as a Sensitive Nephelometer

Presented By: JAMES OUIMETTE, Sonoma Ecology Center

The first objective of this presentation is to show that the Plantower PMS5003 sensors in the PurpleAir monitors can be used to measure and predict the 1-h average submicron aerosol light scattering coefficient nearly as well as expensive integrating nephelometers. Second, to show that the effectiveness of the PurpleAir in estimating PM2.5is due both to the PMS5003 behaving like an imperfect integrating nephelometer and to the mass scattering efficiency of ambient PM2.5aerosols being roughly constant. Recent research (https://doi.org/10.5194/amt-2021-170 [https://doi.org/10.5194/amt-2021-170]) demonstrated that the PMS5003 is not an optical particle counter and that the particle number concentrations it reports in six size bins are not a meaningful representation of particle size distribution. The PMS5003 first size channel, “ >0.3 um”, is a measure of the scattering coefficient, not number concentration. The PMS5003 is equivalent to a cell-reciprocal nephelometer that uses a 657 nm perpendicularly-polarized light source and integrates light scattering from 18 to 166 degrees. Yearlong field data at NOAA’s Mauna Loa Observatory and Boulder Table Mountain sites indicated a strong linear relationship between the 1-h average of the “>0.3 μm” size channel and the submicrometer aerosol scattering coefficient bsp1 measured by TSI 3563 integrating nephelometers. The relationship was bsp1 = 0.015 × “>0.3 μm” (R2 = 0.97) from 0.4 Mm-1to 500 Mm-1. This corresponds to 1-h average submicrometer aerosol mass concentrations ranging from approximately 0.2 to 200 µg m-3. A physical-optical model of the PMS5003 was developed to estimate scattered light intensity on the photodiode. Field data showed that 1-h average “>0.3 μm” was linearly proportional to the model-predicted scattered light intensity over 4 orders of magnitude. The presentation will show that model predictions were consistent with experimental data on the PMS5003 response to a variety of aerosols and relative humidity. *Selected for a Lighting Talk

Additional Authors: JAMES OUIMETTE, Jessica Tryner, John Volckens

In-Person Poster Display: #104

Development of a Wearable Monitor for Characterizing Personal Exposures to Particulate Matter and Volatile Organic Compounds

Presented By: Jessica Tryner, Colorado State University

Exposures to air pollutants, including particulate matter (PM) and volatile organic compounds (VOCs), pose health risks in both occupational and everyday settings. Limited data on personal PM and VOC exposures undercut the evidence-base needed to develop exposure standards and/or reduce risks, while the cost and logistics of outfitting individuals with wearable sampling equipment present barriers to large-scale exposure assessment. We developed a small, quiet, wearable monitor with the goal of making it easy to deploy many monitors to measure personal exposures to PM and VOCs across many individuals at once. This monitor samples PM onto a filter and VOCs onto a thermal desorption tube. The monitor also includes a GPS, as well as low-cost real-time sensors for both PM and total VOCs, to capture spatiotemporal variations in exposures. The monitor is 150 × 45 × 38 mm, weighs 200 g, and has a battery life ranging from 14 to 25 h (depending on the operational state of the GPS and real-time PM sensor). Results from laboratory experiments indicate that time-averaged fine particulate matter (PM2.5) and benzene concentrations derived from samples collected using the new monitor are within 10% and 20%, respectively, of those derived from samples collected concurrently using conventional personal sampling equipment. Our pilot field sampling efforts focus on using the new monitors to characterize exposures to PM2.5 and VOCs among workers performing a variety of tasks at an agricultural research facility. *Selected for a Lighting Talk

Additional Authors: Jessica Tryner, Casey Quinn, Emilio Molina Rueda, Christian L'Orange, Ellison Carter, John Volckens

In-Person Poster Display: #105

Autonomous Low-Cost Ozone Sensors: Development, Calibration, and Application to Study Urban-Rural Gradients

Presented By: Shantanu Jathar, Colorado State University

Ozone (O3), a criteria pollutant and atmospheric oxidant, is not routinely measured in rural and remote environments and hence exposure to ozone pollution in these regions remains poorly understood. In this work, we built, calibrated, and deployed five low-cost, autonomous ozone samplers (called MOOS) in northern Colorado, a region that is non-compliant for ozone during the summertime. The autonomous sampler included the following components: (i) an Aeroqual SM50, a heated metal oxide ozone sensor, mounted inside a custom radiation shield, (ii) a power system that consisted of a 30 W solar panel, 108 Wh lithium-ion battery, and charge controller, (iii) a Particle Boron to acquire, process, and transmit data to the Cloud, and (iv) an environmental sensor to measure temperature and relative humidity. In a three-week long colocated study, we found that all MOOS, calibrated to a single day of reference data, compared well against reference monitors with a measurement error between 4-6 ppbv. Manufacturer and laboratory-based calibrations performed much poorly and over- and under-estimated ozone levels at higher and lower ozone mixing ratios, respectively. When deployed in northern Colorado for an additional three weeks to measure and study the east-west gradient, we found that MOOS calibrations based on the colocated study did as well or better than a calibration based on a single day of reference data in the field. Compared to the colocated study, the field study resulted in larger measurement errors for all five MOOS (<9 ppbv). Furthermore, there was modest variability in the field performance across the different MOOS (4 to 9 ppbv) that could not be explained by environmental differences between the different sites (e.g., proximity of the MOOS to the reference monitor, land use type, temperature). Overall, our study indicates that the MOOS shows promise in being used to supplement routine ambient monitoring and characterizing regional ozone pollution. *Selected for a Lighting Talk

Additional Authors: Dylan Giardina, Ilana Pollack, Emily Fischer, Sheryl Magzamen, Jessica Tryner, John Volckens, Shantanu Jathar

In-Person Poster Display: #106

Sensor+ - a versatile platform for high-performance, low-cost AQ multisensor systems

Presented By: Tobias Baur, Saarland University, Lab for Measurement Technology

The gas sensor market is constantly evolving, partly due to the still increasing interest in environmental measurements to fulfill the legislative guidelines or the United Nations sustainable development goals, limiting the exposure of people to certain gases, but also in the context of IoT devices for comfort. New, often digital sensors are rapidly introduced in the market, so that sensors are often replaced by newer versions within a short timeframe. Designing electronics to put these rapidly developing sensors into operation is time consuming and an obstacle to fast evaluation of new sensor systems that make full use of the sensors´ capabilities. Therefore, a hard- and software platform for multisensor systems, especially metal oxide semiconductor (MOS) gas sensors, but also miniaturized photoacoustic CO2 sensors or digital environmental sensors (p, T, RH) is presented. The goal was to create a hardware system offering a suitable compromise between easy installation and commissioning of various sensors and useful functionalities for sensor signal capturing also allowing simple optimization of their operating mode for different applications. Different versions of the system allow the evaluation of a variety of analog and digital MOS sensors and to control temperature cycled operation (TCO) to increase selectivity, sensitivity, and stability. The presentation of the platform will discuss the three categories electronic hardware, firmware, and control software. The system can be combined with modules for wireless data transfer or local data storage on SD cards as well as different power supplies to also allow mobile applications. The platform is therefore suitable for research as well as outreach activities but can also be directly implemented as a component in application specific sensing solutions. *Selected for a Lighting Talk

Additional Authors: Tobias Baur, Christian Fuchs, Christian Bur, Andreas Schütze

In-Person Poster Display: #107

Merging sensor data with other air pollution data sources: methods and benefits

Field Testing PurpleAir Sensors Alongside IMPROVE Air Quality Monitoring System

Presented By: Hanyang Li, Air Quality Research Center, University of California Davis

Particulate matter (PM) monitoring has developed into a data intensive domain, where large amounts of data can be collected from low-cost sensors (LCS) and ground-based monitoring networks. LCS provides measurements at fine resolution but is low accuracy. Gravimetric analysis of filters used in the routine monitoring networks is more accurate but only yields integrated measurements at low time resolution. Combining LCS with monitoring networks provides a unique opportunity to constrain the uncertainties of LCS, which enables better understanding of air pollution at finer scales. The IMPROVE network has installed PurpleAir PA-II sensors at seven monitoring stations (FRES, GRSM, MACA, SHEN, THRO, WIMO, and MEAD) since 2019. In this study, we evaluated the data quality of PurpleAir sensors at these stations and developed a correction model leveraging network chemical speciation to reduce the discrepancies between PurpleAir PM and IMPROVE gravimetric mass results. Our results suggest the performance of PA-II sensors varies across sites. The raw PM measurements reported by the two PA-II channels agree well at SHEN, GRSM, and THRO, but exhibit systematic biases at the other sites. By examining the measured particle number concentrations, the discrepancies between the two channels are associated with the 1.0 µm/dlbin. In addition, the PA-II overpredicts IMPROVE PM2.5 at most sites (with factors of 1.2 to 2.75, R2of 0.6, and MSE of 10), excepting MEAD where PA-II underpredicts by 40 %. By applying the EPA PurpleAir correction to our data, the over-estimation of PA-II is reduced but the uncertainty across different locations remains. Our proposed model with corrections of chemical species effectively improves the precision of PurpleAir, resulting in a slope of 0.97, R2of 0.95, and MSE of 1.

Additional Authors: Hanyang Li, Nicholas Spada

In-Person Poster Display: #87

Publicly available low-cost sensor measurements for PM2.5 exposure modeling: Guidance for monitor deployment and data selection

Presented By: Jianzhao Bi, University of Washington

High-resolution exposure modeling is critical for assessing the health effects of ambient PM2.5 in epidemiological studies. Using sparse regulatory PM2.5 measurements as principal model inputs may affect the models’ accuracy in predicting PM2.5 spatial distribution. In this study, we used the PM2.5 measurements from a publicly available commercial low-cost PM2.5 network, PurpleAir, with an external validation dataset at the residential locations of a representative sample of participants from the ACT-AP study, to improve the accuracy of exposure prediction at the cohort participant locations. We also proposed a metric based on principal component analysis - the PCA distance - to assess the similarity between monitor and cohort locations to guide monitor deployment and data selection. The analysis was based on a spatiotemporal modeling framework with 51 “gold-standard” monitors and 58 PurpleAir monitors for model development, as well as 105 home monitors at the cohort locations for model validation in the Puget Sound region of Washington. After including calibrated PurpleAir measurements as part of the dependent variable, the external spatiotemporal validation R2 and RMSE for two-week concentration averages improved from 0.84 and 2.22 μg/m3 to 0.92 and 1.63 μg/m3, respectively. The external spatial validation R2 and RMSE for long-term averages over the modeling period improved from 0.72 and 1.01 μg/m3 to 0.79 and 0.88 μg/m3, respectively. The PurpleAir monitors with shorter PCA distances improved the model’s prediction accuracy more substantially than the monitors with longer PCA distances, supporting the use of this similarity metric.

Additional Authors: Jianzhao Bi, Nancy Carmona, Magali Blanco, Amanda Gassett, Edmund Seto, Adam Szpiro, Timothy Larson, Paul Sampson, Joel Kaufman, Lianne Sheppard

In-Person Poster Display: #88

Citizen-Driven Air Sensing Network to Study Intra-Urban Heat (UHI) and Pollution Island (UPI)

Presented By: LU LIANG, University of North Texas

Defined by higher ambient air pollution in urban areas as compared to the non-urban counterparts, the urban pollution island (UPI) effect presents a critical socio-environmental and health issue to urban dwellings. Dallas-Fort Worth has been witnessed as one of the fastest-growing metroplexes and its air quality has been consistently ranked among the worst in the U.S. However, whether the accompanied significant urbanization has resulted in an overall increase in the UPI remains unknown. This project will utilize low-cost sensors to build an urban observation network in a growing county of DFW, for enhanced capacity of real-time, near-surface, and representative PM2.5 monitoring. We have adopted a stratified sampling design modified from the New York City Community Air Survey to place the 100 PurpleAir sensors in six sampling strata. A variety of advanced machine learning algorithms are tested to calibrate the sensor by collocating the PA sensors with a reference instrument. Using the calibrated PM2.5 data, we will model daily PM2.5 concentration at the 30-m resolution through the integration of ground sensor measurements with satellite observations. The high spatial-temporal resolution urban pollution mapping results will shed light on revealing places and times when people will be more exposed to air pollution, with further implications for health and other socio-economic outcomes.*Selected for a Lighting Talk

Additional Authors: LU LIANG

In-Person Poster Display: #89

Evaluation of Methods for Using Low-cost Sensors to Improve Hourly Air Quality Forecasts for the United States

Presented By: Nathan Pavlovic, Sonoma Technology, Inc.

Air quality forecasts are used to support air quality management programs and help the public make decisions to protect their health. Improving the accuracy and resolution of air quality forecasts are important objectives to provide more actionable information to these stakeholders. Data from high-density low-cost sensor networks present an opportunity to greatly improve the spatial fidelity of forecasts, especially when they are used with air quality model data and other key data such as transportation and wildfire emissions. In this work, we prototyped three statistical methods for spatial air quality forecasting of PM2.5 at an hourly resolution. Each prototype incorporates the same input data, including reference-grade and low-cost sensors, and operational forecast chemical transport model results. Using the prototypes and independent reference-grade observations, we perform a head-to-head evaluation of the forecast characteristics of each method, including 10-fold cross validation accuracy, bias, computational requirements, hour-by-hour continuity, and interpretability. In addition, we assess the performance improvement from the inclusion of low-cost sensors by removing the sensor observations from the forecast input. The results provide a clear comparison of the strengths and weaknesses of each method, and can be used to guide the development of operational, high-resolution forecasting products.

Additional Authors: Nathan Pavlovic, Anondo Mukherjee, David Miller, Jennifer DeWinter

In-Person Poster Display: #92

Mobile Monitoring/Monitoring Mobile Sources

Next-Generation Heavy-Duty Vehicle Enforcement with Roadside Emissions Monitoring Devices

Presented By: Hang Liu, California Air Resources Board

Heavy-duty (HD) vehicles are a significant source of emissions that contribute to adverse health outcomes. The California Air Resources Board (CARB) has adopted and implemented multiple regulations over the years to reduce emissions from HD vehicles. Previous studies have shown that a small percentage of HD vehicles are responsible for a disproportionate portion of total HD emissions. With more than one million trucks operating annually in California, it is important to be able to efficiently identify high-emitting vehicles to target enforcement efforts, especially in communities heavily impacted by truck traffic. CARB’s Enforcement Division has been developing and deploying the Portable Emissions AcQuisition System (PEAQS) roadside emissions monitoring devices as a support tool for current and future heavy duty diesel enforcement programs. PEAQS devices estimate vehicle emissions by capturing truck exhaust plumes and utilizing sensors to measure black carbon, oxides of nitrogen (NOx), and carbon dioxide for identifying high emitting vehicles. Two devices have been deployed since August 2019, with each generating millions of second-by-second emission records and tens of thousands of vehicle license plate records from the integrated automatic license plate reader (ALPR) systems each month. CARB’s Enforcement Division plans to deploy more devices to establish a monitoring network within the State. This presentation will show how this large dataset from a network of PEAQS units can be used to draw insights on HD traffic patterns, vehicle composition, and vehicle emission characteristics. It will also demonstrate how CARB has been developing decision support systems to process, analyze, and visualize PEAQS and other data to support CARB’s next-generation data-driven enforcement program.

Additional Authors: Hang Liu, Cody Howard, Walter Ham

In-Person Poster Display: #93

Mobile Monitoring on Trash Trucks Using Sensors and Drive-By Calibrations with Reference Grade Monitors

Presented By: Jessa Ellenburg, 2B Technologies

In collaboration with the City and County of Denver, and as part of a grant from the National Institutes of Health, the concept of Park-By and Drive-By calibrations of mobile sensors was evaluated. Five sensor suites (PAMs) measuring PM2.5, CO, NO2 and CO2 were installed on five City of Denver trash trucks in August of 2021. Additionally, two reference grade monitoring packages (AQSyncs) were installed where the trash trucks park and one was installed at a transfer station the trucks visit. The goals of this are study are to determine: -Feasibility of a co-located calibration to keep sensor data quality high -Recommended duration of the co-location -Recommended frequency of the co-location -Calibration algorithms for data correction in the cloud -Recommended frequency of data calibration adjustments This study will help determine the feasibility of using city fleet vehicles such as trash trucks to map air pollutants through the city. Potential applications and benefits of such programs from a city perspective will also be presented.

Additional Authors: Jessa Ellenburg, Albert Presto, John Birks, Craig Williford, Matt Beach, Michael Ogletree, Donnie Cruz

In-Person Poster Display: #94

Performance targets for air quality sensors

Analysis and Observations from the Deployment and Testing of a Commercially-Available VOC Sensor

Presented By: Ashley Collier-Oxandale, South Coast Air Quality Management District

Interest in measuring air toxics and hazardous air pollutants, particularly by communities and members of the public, has led to an increased interest in commercially-available VOC sensors. However, between the range of compounds that sensors may encounter and the different operating principles these sensors rely on, there is substantial complexity to using these devices for applications such as ambient monitoring. For example, metal-oxide (MOx) semi-conductor sensors vary significantly in their sensitivity to different compounds and their susceptibility to interferents. These characteristics impact their ability to provide absolute TVOC (total volatile organic compounds) or TNMHC (total non-methane hydrocarbons) concentration values. We will present an in-depth exploratory analysis of data collected using new and aged commercially-available MOx-based VOC sensors (less than USD $300 per sensor). Data was collected through a field co-location with a reference instrument and laboratory tests. Analysis of Variance (ANOVA) and regression analysis were used to explore the "off-the-shelf" performance of the sensors and better understand key limitations. For example, field and lab data analysis identified and confirmed cross-sensitivities to temperature and ozone, which are especially important if the intended application is ambient monitoring. In addition, we will discuss how this work helped shape the new testing protocols for VOC sensors in development by the South Coast AQMD's Air Quality Sensor Performance Evaluation Center (AQ-SPEC), ensuring the testing is appropriate and useful given the range of available VOC sensor technologies. More broadly, this initial deployment and testing have assured that the results of AQ-SPEC VOC sensor evaluations will provide the detail and nuance necessary to inform the public and translate into practical considerations for the use of VOC sensors. *Selected for a Lighting Talk

Additional Authors: Ashley Collier-Oxandale, Xiaobi (Michelle) Kuang, Wilton Mui, Randy Lam, Jason Low, Vasileios Papapostolou, Andrea Polidori

In-Person Poster Display: #65

Particle Sizing Performance Evaluation of Low-Cost Particulate Matter Sensors

Presented By: Emilio Molina Rueda, Colorado State University

Outputs from low-cost particulate matter sensors typically include mass and number concentrations reported across multiple size bins. However, particle sizing is challenging even for laboratory-grade instruments, especially those that rely on light scattering. Under ideal conditions, discrete pulses would be obtained for individual particles. Mapping each pulse to a size bin, based on amplitude and width, would be a relatively simple classification problem, although other factors (e.g., scattering coefficients) would introduce uncertainty. Real conditions, in contrast, lead to ambiguous, noisy signals that can’t be interpreted easily. Pulse overlaps due to multiple particles in the detector and sampling errors affect the signal in different ways. If coincidence errors (i.e., multiple particles seen as a single “larger” particle) are sporadic, size and count errors are introduced. If pulse overlap is widespread, particle sizing becomes more complex because identifying different particle sizes in a combined signal is not trivial. Although data reported by some low-cost sensors are well-correlated with mass concentration changes, size distributions range from having no basis in measurement data to vaguely representative. We evaluated the particle sizing performance of multiple sensors including the Sensirion SPS30 and the Piera IPS-7100. We conducted laboratory experiments with monodisperse, polydisperse, and bi-modal aerosols, in addition to an ambient monitoring test to explore whether sensor outputs are estimated from actual particle sizes and counts. Preliminary results from ambient monitoring show that counts of particles <1 µm reported by the IPS-7100 and SPS30 are highly correlated with data reported by a collocated GRIMM EDM 180 monitor (r = .92, p < .001 and r = .94, p < .001 respectively). Both sensors had significantly lower correlation coefficients for particles in the 1 to 2.5 µm range (r = .54, p < .001 and r = .27, p < .001), and for larger particles. *Selected for a Lighting Talk

Additional Authors: Emilio Molina Rueda, Jessica Tryner, Cassey Quinn, Christian L'Orange, Ellison Carter, John Volckens

In-Person Poster Display: #63

Evaluation of AQY1 Sensors and Implications for Community-led Projects

Presented By: Joshua Stratton, Rider University

While access to lower-cost sensors has greatly increased over the last decade, questions remain about their reliability for community-led projects. This study aimed to compareAeroqual’s AQY1 sensors (n=6) with state operated real time ozone (O3) (ultraviolet absorption), nitrogen dioxide (NO2) (chemiluminescence) and fine particulate matter (PM2.5) monitors (beta-attenuation) under real-world conditions in New Jersey (USA). Aeroqual AQY1 sensors were evaluated to understand their potential for community-led hotspot monitoring projects utilizing high-time-resolution data (<60 min averages). Undergraduate students operated, maintained, and analyzed the data from these sensors as part of a larger project to understand potential pollutant hotspots around campus. During collocation, AQY1 O3 concentrations were correlated (R2= 0.62-0.82) with the reference monitor concentrations (monitor mean =33 ppb and maximum = 80 ppb) but were largely underestimated (slope = 0.22-0.29). NO2 concentrations were low (monitor mean = 8 ppb and maximum = 25 ppb) with little correlation between sensors and the reference monitor. NO2 measurements were found to exhibit a dependence on relative humidity and a shifting response over the study period. During collocation, AQY1 PM2.5 concentrations agreed with the reference monitor when including impacts of fireworks during the 4th of July celebrations (R2= 0.83-0.85 and slope = 0.99-1.09), however, it decreased when excluding this episode (slope = 0.42-0.50, R2 = 0.40-0.51). A calibration from this study period corrected for large underestimations of O3, but NO2 remained inaccurate.O3 and PM2.5measurements show promise for community-led projects based on the accuracy following correction at elevated concentrations. However, data correction, harmonization, and interpretation will likely require some significant amount of effort from community groups likely necessitating partnerships with local air monitoring agencies and EPA regional offices. *Selected for a Lighting Talk

Additional Authors: Joshua Stratton, Jessica Barone, Nicolas Kaltenhauser, Daniel Druckenbrod, Rachael Leta-Graham, Luis Lim, Olga Boyko, Jessica Munyan, Karoline Barkjohn, Samuel Frederick, Andrea Clements

In-Person Poster Display: #61

Sensor Toolkit: A Python Library for Standardizing the Ingestion, Analysis, and Reporting of Air Sensor Data for Performance Evaluations

Presented By: Samuel Frederick, Oak Ridge Associated Universities

Past efforts to establish open-source software tools for reporting sensor performance have been limited. Sensor and reference data formats vary widely, and in turn, many existing software packages evaluate only one type of sensor model. Other packages allow broader utilization of air quality data yet may not be specifically tailored for evaluating sensor performance against reference data. Additionally, these packages do not provide means for summarizing sensor performance in a reporting template using common statistical metrics and figures. To encourage broader utilization of the U.S. Environmental Protection Agency’s (EPA) recommended performance metrics and target values for sensors measuring fine particulate matter (PM2.5) and ozone (O3), EPA developed a new, open-source Python library named “sensortoolkit”. The library compares collocated sensor data against reference monitor data and includes methodology to re-format both datasets into a standardized format using an interactive setup module. Library modules are included for calculating EPA’s recommended sensor performance metrics and for making relevant plots. These metrics and plots can be used to better understand sensor accuracy, precision between sensors of the same make and model, and the influence of meteorological parameters at 1-hr and 24-hr averages. Results can be compiled into the reporting template included alongside EPA’s performance targets documents. The sensortoolkit library is designed for any user, from novices to researchers. *Selected for a Lighting Talk

Additional Authors: Samuel Frederick, Karoline Barkjohn, Rachelle Duvall, Andrea Clements

In-Person Poster Display: #69

The Modified Target Diagram for the performance evaluation of low-cost sensors

Presented By: Sinan Yatkin, Joint Research Centre

The modified Target Diagram (MTD) was developed to evaluate the performance of low-cost sensors (LCS) in comparison with reference methods by reporting relative expanded uncertainty and its’ contributors. MTD includes several information, e.g., compliance with the European regulation, sources of error and how to diminish them, misdefinition of LCS calibration model etc. A direct glance at MTD shows lack of sensitivity of LCS, over or underestimation of LCS and the major contributor(s) to uncertainty being either random errors or bias. It allows user to assess the effect of selecting different regression types (LCS versus reference data) and residual fitting on the LCS expanded uncertainty. The ordinary least squared regression with fitted residuals and variable between reference analyser uncertainty rather than constant ones yielded more realistic LCS uncertainty compared to the other options. MTD is fast visual tool to extract several information on evaluation of any candidate method against reference method. *Selected for a Lighting Talk

Additional Authors: Sinan Yatkin, Michel Gerboles, Annette Borowiak, Silvije Davila, Laurent Spinelle, Alena Bartolova, Frank Rene Dauge, Philipp Schneider, Martine Van Poppel, Jan Peters, Christina Matheeussen, Marco Signorini

In-Person Poster Display: #70

Development of Comprehensive Test Protocols for the Performance Evaluation of VOC Air Quality Sensors

Presented By: Wilton Mui, South Coast Air Quality Management District

Conventional VOC monitoring relies on equipment and methods that are time consuming, labor intensive and expensive to operate. With recent advancements in sensor technology, the emergence of VOC sensors can provide a higher time and spatial resolution for air quality and fence-line VOC monitoring. Understanding the operation and function of VOC sensors, as well as their selectivity and sensitivity to various VOC compounds, requires systematic evaluation of VOC sensors. Recognizing the urgent need for protocols to systematically evaluate VOC sensors, the South Coast AQMD Air Quality Sensor Performance Evaluation Center (AQ-SPEC) has taken the initiative to develop testing protocols to evaluate VOC sensors for their ability to measure a wide range of VOC concentrations, perform in the presence of different environmental and gaseous interferents, and characterize their behavior in both ambient and laboratory simulated outdoor environments. The performance of VOC sensors was evaluated using metrics such as data recovery, intra-sensor variability, accuracy, and correlation to reference instruments. ANOVA was also performed to determine important explanatory variables that drive sensor response.

Additional Authors: Xiaobi (Michelle) Kuang, Wilton Mui, Hang Zhang, Ashley Collier-Oxandale, Vasileios Papapostolou, Andrea Polidori

In-Person Poster Display: #64

Performance Evaluation frameworks for low cost air quality sensors.

Presented By: Jalal Awan, Pardee RAND Graduate School

Typically, the performance of low cost sensing devices is assessed using the mean error or correlation coefficients [https://www.sciencedirect.com/topics/earth-and-planetary-sciences/correlation-coefficient] (R2) values withrespect to laboratory equipment. Even though these criteria are good starting points for selection of low-cost sensors, they do not reflect various aspects of sensor performance (accuracy, consistency and reliability) under different geographic locations, climate conditions, pollution concentrations, and more importantly, changes in performance over time (Chang, Zhang, & Jiang, 2019). Based on our findings from deployment of ~25 low cost air quality sensors across the 5 zip codes in the City of Santa Monica and 20 sensors in Pittsburgh, we propose a maintenance and performance evaluation regime for low cost sensors based on lessons learnt from our experiment (https://www.prgs.edu/news/2019/cazier-air-quality-analysis.html [https://www.prgs.edu/news/2019/cazier-air-quality-analysis.html]). First, we conduct a review of literature to identify and shortlist best practices for preventive and corrective maintenance of remote sensing devices through Google Search and WoK. Based on findings from the literature review, we conducted an overall performance evaluation of our deployed sensors through data analysis and a qualitative survey of sensor users. Finally, we propose a user-friendly maintenance and re-calibration regime (to correct for drift) for a typical low-cost air quality sensing device. Throughout the analysis, our 'gold standard' for baseline measurements is the nearest EPA FRM / FEM sensing unit. My preliminary findings indicate the need for more robust sensors, particularly for regulatory sensing devices under community air monitoring paradigms such as those under California AB32. Finally, I present recommendations for state, local and federal environmental monitoring agencies.

Additional Authors: Jalal Awan

In-Person Poster Display: #373

How to get accurate measurements from micro-sensors?

Presented By: Aymerick Prunes, Application Engineer, ECOMESURE

The covid-19 crisis has accelerated public awareness about air pollution. The deployment of massive quantities of low-cost sensors is already ongoing in schools (CO2 sensors) and cities (PM2.5 sensors) even though most sensors are lacking accreditation/certification. Meanwhile new monitoring stations manufacturers are also appearing on the market, some with little to no expertise in air quality monitoring. How accurate are these modern technologies? How to ensure the sensors provide precise and meaningful data? The accuracy of the sensors depends on two key factors: the technology used (electrochemical, optical, PID, infrared, MOx...) and the calibration method and verification process applied to guarantee the quality of the values measured. Thanks to its 30 years’ experience in metrology and maintenance of air quality sensing material, Ecomesure has developed a disruptive technology to improve the calibration process of its devices using Artificial Intelligence. The process involves an automated multi-point calibration followed by a cross-comparison in an ISO-certified laboratory between calibrated sensors and reference devices data. Side-by-side comparison tests between calibrated sensors are performed to ensure true repeatability. The combination of these two methods ensures the quality of measurement at the factory and certifies the quality of the data measured by the sensors. In the field intercomparison with Air quality monitoring reference stations can also be carried depending on the project and expected accuracy of measurements in order to adjust the sensor offset (e.g. Canton of Geneva, Port Autonome de Strasbourg, ENGIE etc.). Technological breakthroughs make it possible to maximize these operations and further improve the reliability of measurements by integrating artificial intelligence models at the time of calibration. This is the object of a new patent that Ecomesure has filed and will present in the event.

Additional Authors: Julie PELLETIER, Benjamin Vancoillie, Aymerick Prunes, Damien PELLETIER

In-Person Poster Display: #66

Impact of different particle sources on the measurement of PM by low-cost sensors.

Presented By: Laurent Spinelle, Ineris

During the last years, as a member of the French National Reference Laboratory for monitoring air quality (LCSQA), Ineris has been developing and using "a PM enhanced generation system for inter-laboratory comparisons of automatic PM analysers". This facility is based on particle generation by nebulising a mixture of ammonium sulphate and ammonium nitrate diluted in distilled water using a TSI model 3076 nebuliser. However, this method of generation only produces particles with a distribution size limited to 2.5µm, excluding the possibility of evaluating the performance of measurements for the sole larger particles, such as those contributing to the PM10 mass. The objective of the work presented was to evaluate the possibility of generating particles with a larger granulometry and their impact on the measurement carry out by PM low-cost sensor in comparison with an automatic measurement system. This work has been divided in two parts: - We first studied the impact of the type of generation method used on the mass concentration and granulometry of the particles generated using the same saline solution. We have shown that whatever the nebuliser used, i.e. whatever the limit imposed by the cut-off diameter, the generation from saline solution produces particles with a size centred around 600nm and corresponding mainly to PM2.5. - In a second step, we studied the possibility to generate large particles by using a dispersion of Arizona dust in water. Indeed, even if the particle size distribution remains centred around 600nm, the presence of larger diameter particles has a major impact on the PM10 fraction. The simultaneous use of both methods showed the possibility of adding an independent contribution for the PM10 fraction. Finally, we propose to discuss the impact of the two generation methods, independently and combined, on the measurements of a PM low-cost sensor in comparison to an automatic measurement system commonly used for the air quality monitoring.

Additional Authors: Laurent Spinelle, Jason Bardou, Brice Berthelot

In-Person Poster Display: #67

Calibrating low-cost sensors for wildfire smoke: how algorithmic corrections to low-cost sensor data can help meet USEPA and other performance targets

Presented By: Levi Stanton, Clarity Movement Co.

With the proliferation of low-cost sensor technology and the consequent availability of expanded air quality datasets, calibrated low-cost sensor performance is improving every year. Clarity regularly generates seasonally and regionally-specific calibration models to bring sensor performance in line with data quality guidelines — such as the recently released United States Environmental Protection Agency (USEPA) PM2.5 Performance Targets. Anticipating a long and challenging wildfire season in 2021, we developed a new and improved PM2.5 calibration model to account for elevated particulate matter air pollution in Western North America. This presentation will describe the approach we took in developing this new model and provide performance metrics for the model, which represent a substantial improvement in sensor performance during periods of elevated ambient particulate matter. Using collocated Clarity Node-S devices at 13 reference sites across five states (California, Idaho, Montana, Oregon, and Washington) we generated a dataset of collocated particulate matter measurements for the period from July 2019 to July 2021. We separated the dataset into model training and model testing datasets. One location (in Montana) was kept out of the training dataset to serve as a completely independent site and to test how well the model performs in different geographies and climates. We evaluate the performance of the model using relevant USEPA PM2.5 Performance Targets. While the uncalibrated PM2.5 measurements do not meet all USEPA performance targets, the calibrated PM2.5 data (both hourly and 24-hour) met all of the recommended targets for data quality.

Additional Authors: Levi Stanton

In-Person Poster Display: #68

Sensor Networks: From nuts and bolts to real-world impacts

Measuring the Spatial and Temporal Variations of Air Pollution in Complex Urban Environments: Results from the Richmond Air Monitoring Network

Presented By: Boris Lukanov, Physicians, Scientists and Engineers for Healthy Energy (PSE)

Low-cost sensor networks are an attractive option for characterizing spatially heterogeneous pollutant concentrations in urban environments and detecting air pollution episodes at locations not covered by government-operated air monitoring sites. Recent developments in instrumentation, communication, and data analysis have enabled the deployment of such distributed sensor networks in support of both regulatory activities and scientific research/citizen science. The Richmond Air Monitoring Network is a dense network of 50 Aeroqual AQY air monitors collecting PM2.5, NO2and O3measurements every minute of the day across various locations in Richmond, California. Sites were selected through a community outreach process and span a wide range of urban land-use characteristics, including industrial, commercial, residential, near-highway, near local and arterial roads, near traffic intersections, and others. We present initial analyses of data collected by the network over a two-year study period. Sensor measurements were used to determine the spatial variability of air pollution over various timescales, including hourly, weekly, and seasonally. A breakdown of air pollutant concentrations by time of day reveals general daily trends for the three different air pollutants by neighborhood and land use area. Hotspots and neighborhood-level air pollution episodes were identified and compared with data from the small number of regulatory sites in the region. Strong fluctuations in air pollution were periodically observed over hourly, diurnal, and weekly cycles, reflecting the effects of localized traffic and industrial facilities in the area. The results demonstrate how a dense network of community air monitors can be used to reveal the complex spatiotemporal dynamics of air pollution within urban neighborhoods. The findings also illustrate how a distributed sensor network can be useful for addressing current limitations in the spatial coverage of government air monitoring.*Selected for a Lighting Talk

Additional Authors: Boris Lukanov, Audrey Smith, Karan Shetty

Virtual Poster Discussion: Thursday, May 12, 8:00 a.m. PT

In-Person Poster Display: #24

How accurate are out-of-the-box low-cost particulate matter sensor systems?

Presented By: Nuria Castell, NILU-Norwegian Institute for Air Research

Increased availability of commercially-available low-cost air quality sensors combined with increased interest in their use by citizen scientists, community groups, and professionals is resulting in a rapid adoption despite data quality concerns. We have characterized three out-the-box PM sensor systems under different environmental conditions using field colocation against reference equipment. The sensor systems integrate Plantower 5003, Sensirion SPS30 and Alphasense OCP-N3 PM sensors. The first two use photometry as a measuring technique, while the third one is an optical particle counter. For the performance evaluation we co-located 3 units of each manufacturer and compared the results against optical (FIDAS) and gravimetric (KFG) methods for a period of 7 weeks (28 August to 19 October 2020). During the period from 2nd and 5th October, unusually high PM concentrations were observed due to a long-range transport episode. The results show that the highest correlations between the sensor systems and the optical reference are observed for PM1, with coefficients of determination above 0.9, followed by PM2.5. All the sensor units struggle to correctly measure PM10, and the coefficients of determination vary between 0.45 and 0.64. This behaviour is also corroborated when using the gravimetric method, where correlations are significantly higher for PM2.5 than for PM10. During the long range transport event the performance of the photometric sensors was heavily affected, and PM10 was largely underestimated. The sensors also showed a decrease in accuracy when the ambient size distribution was different from the one the manufacturer had calibrated the sensor, and during weather conditions with high relative humidity. When interpreting and communicating air quality data measured using low-cost sensor systems it is important to consider such limitations in order not to risk misinterpretation of the resulting data.

Additional Authors: Nuria Castell, Matthias Vogt, Philipp Schneider

In-Person Poster Display: #17

Technical Challenges and Solutions to Improve the Accuracy of Low-Cost Sensor Data

Presented By: Christi Chester Schroeder, IQAir North America

Low-cost PM sensors hold great promise in supplementing reference-grade instruments in providing hyper-local and hyper-responsive air quality information. The availability of this low cost technology has resulted in the expansion of ambient air quality monitoring networks around the globe by providing more economically feasible options empowering organizations and individuals such as NGOs, community activist groups, and educational institutions to generate and access publicly available air quality data. However, low-cost sensor networks suffer from a number challenges, many of which can introduce systematic bias and negatively impact overall data quality. These challenges are broad in scope and range from improper sensor placement to more technical obstacles such as the inability to distinguish between PM10 and PM2.5and the overestimation of PM concentrations by optical sensors due to humidity and fog. Comprehensive quality assurance protocols are necessary to ensure data integrity throughout the entire process of data acquisition, data cloud transmission, data processing, and data reporting. The presentation will provide a systematic assessment of factors impacting low-cost sensor data quality through each step of the low-cost sensor data pipeline, a discussion of solutions we have implemented to mitigate sources of bias, and the challenges we are still working to address.

Additional Authors: Christi Chester Schroeder

In-Person Poster Display: #27

Evaluation and modeling of data from low-cost air quality sensors for accurate PM2.5 estimation

Presented By: Dinushani Senarathna, Department of Mathematics, Clarkson University

PM2.5 is a critical air quality parameter associated with many air pollution-mediated adverse health effects. In the United States, the Environmental Protection Agency (EPA) provides precise measurements of PM2.5 but a sparse distribution of EPA monitoring sites limits the availability of such data at high spatial resolution. Recent development in low-cost sensors such as PurpleAir (PA) sensors, which can be deployed at high density, show promise to overcome this challenge. However, data generated by these sensors are noisier and tend to overestimate PM2.5 relative to EPA measurements Although models can be used to improve the accuracy of PM2.5 estimates from PA sensor measurements, the validity of these models for sensors in different locations is not fully understood. In this work, we used the PA sensor network and EPA data from Cook County in the Chicago area to evaluate the robustness of correction models and their applicability for sensors in other locations. Our objectives were, (a) to investigate the impact of distance of PA sensor from EPA on model accuracy, and (b) to determine if data from multiple low-cost sensors can be used to generate more precise estimates of PM2.5. Our results demonstrated a dependence of model prediction accuracy on distance, with the accuracy reduced significantly for distances > 30 km from the PA sensor site. Moreover, a higher prediction accuracy was observed with models built using multiple PA sensors (R2 = 0.50 ~ 0.70, RMSE = 2.5 ~ 3.0) than models built with single PA sensor (R2 = 0.30 ~ 0.60, RMSE = 1.9 ~ 3.5), although the improvements were minimal when data from more than 3 sensors were used. Our results indicate that the accuracy of PM2.5 estimation from low-cost sensor data can be improved by considering the distance and incorporating multiple sensors in the model. *Selected for a Lighting Talk

Additional Authors: Dinushani Senarathna, Vijay Kumar, Supraja Gurajala, Shantanu Sur, Suresh Dhaniyala, Sumona Mondal

In-Person Poster Display: #26

Sensor networks to evaluate local air quality impacts from changing traffic scenarios near an elementary school

Presented By: Jelle Hofman, Flemish Institute for Technological Research (VITO)

To improve local air quality, safety and wellbeing near an elementary school, the municipality of Kampenhout (Belgium) implemented traffic restriction measures during school hours. In an attempt to evaluate the resulting air quality impact, two commercially available sensor boxes (Airly and Kunak) were applied for a 3-month period (12/2/2021-14/5/2021) at three locations near the school: two traffic locations (one at school entrance and one outside traffic restriction area) and one background location. Before and after deployment, the sensor performance was evaluated next to an AQMS (Air Quality Monitoring Station). From the co-location measurements, sensors showed low inter-sensor variability and good correlation with reference measurements, but sometimes low accuracy. After a linear slope correction, good sensor performance was achieved for PM1 (R² = 0.82-0.87, MAE = 2.16-2.33 µg m-3), NO (R² = 0.88-0.98, MAE = 1.20-2.77 µg m-3)and NO2(R² = 0.92-0.95, MAE = 3.38-4.67 µg m-3). Different traffic scenarios were tested during deployment, all lasting for about 3 weeks; business as usual (baseline), full road closure (schoolstraat) and single lane closure preventing cut-through traffic (knip), while a holiday period was included as well. Largest absolute concentration differences between the considered locations was observed for NO, followed by NO2 and PM1. From the comparison with simultaneous background measurements, we quantified the local road traffic contribution at the school location in the order NO>>NO2>PM1. After normalizing for the fluctuating background concentration, the implemented traffic scenarios resulted in significant concentration reductions up to 60% for NO and 89% for NO2, while no observable effect was obtained for PM1. This case study, therefore, demonstrates that air quality sensors can quantify local impacts from (even short-termed) traffic measures and can empower local governments to implement new air quality management plans.

Additional Authors: Jelle Hofman, Jan Peters, Martine Van Poppel

In-Person Poster Display: #21

Using Networked Air Quality Sensors to Understand Neighborhood-Scale Differences in Particulate Matter Concentration During Pollution Episodes

Presented By: Kerry Kelly, University of Utah

The rapid proliferation of low-cost air-quality sensors offers great promise for providing highly resolved air-quality measurements. However, data quality remains a challenge, and strategies for assimilating this imperfect data are still being developed. Here, we describe a layered framework that includes: (1) the development of a low-cost PM sensor network that incorporates the University’s own sensors, called AirU, PurpleAir data, and state regulatory measurements; (2) the laboratory and field calibration of the PM sensors; (3) the development and application of event-specific calibration factors; (4) the detection and screening of erroneous sensor data; (5) the strategies for assimilation of the data from more than 200 low-cost and state regulatory sensors using a Gaussian process model to estimate PM2.5 levels and uncertainty dynamically throughout the modeling domain; and (6) the dynamic visualization of the model and uncertainty to communicate PM2.5 levels to the public. The community can view the visualizations through a public-facing website, and they can access the sensor data through an API. Thus far, this network has generated a rich set of PM measurements, capturing several severe PM episodes resulting from persistent cold air pools, fireworks, wildfires, and dust events. The results illustrate dramatic geospatial differences in PM2.5 concentration during some of these episodes that would not have been observed from the regulatory monitors alone. This infrastructure is currently being used by research collaborators studying environmental justice, asthma exacerbations, and simulation of wildfire plumes. The tools sensors and data infrastructure are currently deployed in Salt Lake City, UT; Chattanooga, TN; Cleveland, OH; Kansas City, MO; and Springfield, MA. COI: Drs. Gaillardon, Kelly and Whitaker have a financial interest in Tetrad, LLC, which commercializes environmental monitors. *Selected for a Lighting Talk

Additional Authors: Kerry Kelly, Wei Xing, Tom Becnel, Samy Charas, Anthony Butterfield, Glenn Ricart, Geoff Millener, James Starcev, Krystal Pollitt, Pierre-Emmanuel Gaillardon, Ross Whitaker

In-Person Poster Display: #29

Atlanta Rail and Port Sensor Project: An Air Quality Pilot Study

Presented By: Ryan Brown, US EPA Region 4

The US Environmental Protection Agency (EPA) Region 4 and Georgia Environmental Protection Division (EPD) conducted a pilot air quality study around the largest railyard in Atlanta, Georgia using lower-cost air sensor technology. The objectives of the Atlanta Rail and Port Sensor (RAPS) project were to evaluate air quality near the railyard and to evaluate the utility of air sensor technologies in understanding near-source air quality patterns. Ten PurpleAir particulate matter sensors and an Aethlabs microaethalometer black carbon sensor were used to collect over a year of air quality data at various distances and directions around the railyard. Sensor PM2.5 measurement precision and accuracy were evaluated at an Atlanta area regulatory air monitoring site by collocating the sensors with regulatory PM2.5 measurements before and after deployment near the railyard. After data adjustment using EPA’s national correction equation and removal of data that did not meet project quality assurance indicators, sensor PM2.5 hourly measurement data varied approximately 10 to 20 percent compared sensor to sensor (coefficient of variation) and approximately 25 to 50 percent compared to higher quality, regulatory monitoring measurements (normalized root mean squared error). The collocated sensor metrics showed more variance during the collocation period after deployment primarily due to humidity sensor drift on a few sensors. EPA’s national correction equation includes a humidity factor. Overall, in combination with local meteorological data, sensor measurements were of sufficient quality to detect and quantify some patterns in hourly PM2.5 concentrations near the railyard, due to good sensor to sensor precision. This presentation will also share lessons learned and suggestions for study design, field data collection, sensor data processing, quality assurance screening, and analysis for similar air sensor studies. *Selected for a Lighting Talk

Additional Authors: Ryan Brown, Daniel Garver, Will Carnright, Sara Waterson, Alan Powell, Dale Aspy, Katy Lusky, Corinna Wang, Adam Friedman, Ken Buckley, Karl Armstrong, DeAnna Oser, Andrea Clements

In-Person Poster Display: #30

Beyond Field calibration: Advanced network calibration of multigas and particulate sensors devices through remote source data exploitation

Presented By: Saverio De Vito, ENEA

Although representing the best choice for maximum accuracy of low cost air quality multisensors, field calibration robustness is hampered by incomplete training datasets. Usually, co-location data encompass a limited subset of targets, non targets interferents and environmental conditions envelope resulting in severe loss of accuracy due to seasonality of anthropogenic and natural forcers. Continuous calibration strategies exploiting remote data are promising but the real amount of usable calibration tuples is limited. Most literature approaches rely on the detection of rare conditions in which spatial variability of target concentrations is low as occurring during nightime. This results in network data characterized by extremely low concentrations which only led to baseline correction capability. Here, we present a methodology which try to expand the amount and variance of remote calibration data to the benefit of the overall accuracy. Minute resolution data are filtered to extract and average samples in which lowest 20 percentile of that hour' distribution of WE-AE potential occurs. Averaged data is the coupled with closest regional background station hourly average data to build a new calibration tuple.We preliminary test this approach using a 4 weeks colocation experiment of 30 MONICA devices, a gas and particulate multisensor device devised for both portable and fixed deployments. It is based on a suite of 3 Alphasense A4 sensors targeted to CO, NO2 and O3 plus a Plantower PMS7003 particle sensors. Tested calibration algorithms are multiple linear regression and shallow neural network exploiting response from working (WE) and auxiliary electrodes (AE) of relevant sensor (both target and known non target interferents) and temperature data. Results indicate on that NO2 estimation accuracy of MONICA devices trained with 2 weeks of co-location data, can be consistently overcome (MAE decr. by 11%, R^2 incr. by 6%) by using this advanced network calibration approach. *Selected for a Lighting Talk

Additional Authors:

In-Person Poster Display: #32

Using 5G network to enhance air quality sensing in cities

Presented By: Surya Venkatesh Dhulipala, University of British Columbia

The mass roll-out of 5G infrastructure by major telecommunications companies presents a unique opportunity to collect, calibrate, analyze, and disseminate data from low-cost sensors in real-time. We make use of Rogers Communications’ 5G network at University of British Columbia (UBC) campus to calibrate air quality sensors in real-time, make informed decisions (such as changing traffic patterns etc.), and publish a real-time campus air pollution dashboard. The low-latency and edge computing capabilities of 5G network enable real-time streaming and post-processing of such diverse data streams. To quantify the air quality on UBC campus, we installed a network of 8 low-cost air quality sensors (SENSIT Remote Air Quality Monitoring Platform (RAMP)) equipped with Rogers SIM cards across UBC campus in June 2021 to measure air pollutant concentrations at different traffic intersections. The RAMPs are solar-powered, battery-operated and measure PM2.5, O3, CO, CO2, NO and NO2 every 15 seconds. At the same intersections, we also installed anonymized video cameras to track vehicles, buses, pedestrians, and bicyclists (Numina and Blue City). We present an overview of results from six months’ worth of data collected from our RAMPs and traffic sensors between June 2021 - December 2021. We showcase several applications of air quality sensing using a 5G network. First, we calibrate the RAMPs in real-time and combine calibrated air quality data with traffic data. Second, we use these data streams to simulate how changing traffic patterns in real-time can improve air quality. Third, we showcase a campus pollution dashboard for students and commuters to use as they move around campus.

Additional Authors: Surya Venkatesh Dhulipala, James Hindson, Naomi Zimmerman

In-Person Poster Display: #33

Spatiotemporal analysis of PM2.5 using data from Environmental Protection Agency (EPA) and low-cost sensor networks

Presented By: Vijay Kumar, Department of Mathematics, Clarkson University

PM2.5has been linked to numerous pollution-mediated adverse health effects and their monitoring is key for taking preventative and mitigative measures. Accurate measurements of PM2.5concentrations are available at EPA sites, but such data lacks spatial resolution due to a limited number of monitoring locations. In recent years the deployment of low-cost sensor networks has opened up the possibility of acquiring air quality data at a high spatiotemporal resolution. However, the sensitivity, noise, and accuracy of data acquired by low-cost sensors remain a concern. Here, we studied PM2.5measurements made from EPA and Purple Air (PA) sensor networks in the Chicago area to understand the parameters influencing the performance characteristics of the low-cost sensor network. We decomposed the PM2.5time series data into short-term, seasonal, and long-term components using the Kolmogorov–Zurbenko filter. We then extracted different frequency signals of PM2.5data for each of the filtered components. Our analysis shows that the meteorological variables such as temperature, humidity, wind speed, and wind direction impact the long-term trends of PM2.5 but do not have any influence on the short-term component. Furthermore, the PA sensor data capture long-term variations in the PM2.5data with reasonable accuracy, while the short-term variations (<12 hours) are less clear in comparison to the EPA data. We hypothesize that the low-cost sensor networks may have different detection sensitivity to aerosol from different sources and hence care must be taken in their use for evaluating the impact of air quality mitigation policies.

Additional Authors: Vijay Kumar, S. Dinushani Senarathna, Supraja Gurajala, William Olsen, Shantanu Sur, Suresh Dhaniyala

In-Person Poster Display: #36

Co-kriging with a low-cost sensor network to estimate spatial variation of brake and tire-wear metals and oxidative stress potential in Southern California

Presented By: Jonathan Liu, University of California, Los Angeles

Due to regulations and technological advancements reducing tailpipe emissions, an increasing proportion of emissions arise from brake and tire wear particulate matter (PM). PM from these non-tailpipe sources contains heavy metals capable of generating oxidative stress in the lung. Although important, these particles remain understudied because the high cost of actively collecting filter samples. Improvements in electrical engineering, internet connectivity, and an increased public concern over air pollution have led to a proliferation of dense low-cost air sensor networks such as the PurpleAir monitors, which primarily measure unspeciated fine particulate matter (PM2.5). In this study, we model the concentrations of Ba, Zn, black carbon, reactive oxygen species concentration in the epithelial lining fluid, dithiothreitol (DTT) loss, and OH formation. We use a cokriging approach, incorporating data from the PurpleAir network as a secondary predictor variable and a land-use regression (LUR) as an external drift. For most pollutant species, co-kriging models produced more accurate predictions than an LUR model, which did not incorporate data from the PurpleAir monitors. This finding suggests that low-cost sensors can enhance predictions of pollutants that are costly to measure extensively in the field.

Additional Authors: Amber Kramer, Jonathan Liu, Liqiao Li, Rachel Connolly, Micele Barbato, Yifang Zhu

In-Person Poster Display: #28

Standard, Supplemental and Informational Monitoring

Rapid Assessment of Community Air Quality Using Real-time Mobile Air Monitors

Presented By: Evan Williams, The University of Texas at El Paso

This study evaluates the suitability of assessing air quality in a residential community using air pollution monitoring devices installed in a moving vehicle. Ambient outdoor levels of three criteria pollutants (O3, NO2 and PM) were continuously recorded by three U.S. EPA-certified FEM air pollution monitoring devices installed inside a vehicle traveling repeatedly on the same route in a moderate to low traffic community. Spatio-temporal mobile air quality data were aggregated and compared to data collected at two fixed stations, one permanently operated by a state regulatory agency near a university campus, and another temporarily installed by the research team near a major interstate highway. Hourly mobile and stationary pollution data appeared to agree very well with each other for all three pollutants. The magnitudes and the trends of the mobile PM data (both PM10 and PM2.5) resembled those recorded at the state-operated station. O3 was found to be ubiquitously distributed in the region and presented the best agreement between mobile and fixed-station data. The immediate, complicated photochemical reactions of NOx at tailpipe might have contributed to the incongruity between the mobile and fixed-station data, yet NO2 data appeared to follow a similar trend and peaks. It appears promising that community exposures to transportation related air pollutants can be represented by short-term spatio-temporal measurements using mobile air pollution monitors. Mobile air pollution measurements provide a rapid assessment of the air quality in a community without installing multiple stationary sites. Further research on the adequacy of the mobile air pollution data for exposure and health assessment will greatly enhance the applicability of mobile monitoring in community air quality studies.

Additional Authors: Evan Williams, Mayra Chavez, Leonardo Vazquez, Wen-Whai Li

In-Person Poster Display: #37

Temporal variations of ambient air pollutants and meteorological influences on their concentrations in Tehran during 2012–2017

Presented By: Fatemeh Yousefian, Kashan university of medical sciences

We investigated temporal variations of ambient air pollutants and the infuences of meteorological parameters on their concentrations using a robust method; convergent cross mapping; in Tehran (2012– 2017). Tehran citizens were consistently exposed to annual PM2.5, PM10 and NO2 approximately 3.0–4.5, 3.5–4.5 and 1.5–2.5 times higher than the World Health Organization air quality guideline levels during the period. Except for O3, all air pollutants demonstrated the lowest and highest concentrations in summertime and wintertime, respectively. The highest O3 concentrations were found on weekend (weekend efect), whereas other ambient air pollutants had statistically signifcant (P<0.05) daily variations in which higher concentrations were observed on weekdays compared to weekend (holiday efect). Hourly O3 concentration reached its peak at 3.00 p.m., though other air pollutants displayed two peaks; morning and late night. Approximately 45% to 65% of AQI values were in the subcategory of unhealthy for sensitive groups and PM2.5 was the responsible air pollutant in Tehran. Amongst meteorological factors, temperature was the key infuencing factor for PM2.5 and PM10 concentrations, while nebulosity and solar radiation exerted major infuences on ambient SO2 and O3 concentrations. Additionally, there is a moderate coupling between wind speed and NO2 and CO concentrations.

Additional Authors: Fatemeh Yousefian, Sasan Faridi

In-Person Poster Display: #39

The concentration of BTEX compounds and health risk assessment in municipal solid waste facilities and urban areas

Presented By: Fatemeh Yousefian, Tehran university of medical sciences

In this study, human exposure to benzene, toluene, ethylbenzene, xylenes (BTEX), along with their respective risk assessment is studied in four major units (n = 14-point sources) of the largest municipal solid waste management facilities (MSWF) in Iran. The results were compared with four urban sites in Tehran, the capital of Iran. Workers at the pre-processing unit are exposed to the highest total BTEX (151 μg m− 3 ). In specific, they were exposed to benzene concentrations of 11 μg m− 3. Moreover, the total BTEX (t-BTEX) concentrations measured over the conveyor belt were 198 μg m− 3 at most, followed by trommel (104), and active landfills (43). The mean concentration of ambient t-BTEX in Tehran is 100 μg m− 3. On average, xylenes and toluene have the highest concentrations in both on-site and urban environments, with mean values of 24 and 21, and 41 and 37 μg m− 3, respectively. Even though the non-carcinogenic risk of occupational exposure is negligible, BTEX is likely to increase the chance of carcinogenic risks (1.7E-05) for workers at the pre-processing unit. A definite carcinogenic risk of 1.3E-04, and non-carcinogenic effect, of HI = 1.6 were observed in one urban site. With the exception of the pre-processing unit, the citizens of Tehran had higher exposure to BTEX. Overall, BTEX concentrations in the largest MSWF of Iran remain an issue of public health concern.

Additional Authors: Fatemeh Yousefian

In-Person Poster Display: #38

Air Quality Monitoring in Athletics Stadiums: Can low-cost sensor technologies support guidance for international competitions?

Presented By: Mar Viana, IDAEA-CSIC

The aim of this work was to evaluate the extent to which low-cost sensor technologies can contribute to setting air quality guidelines for international sports events. The ultimate goal was to provide recommendations on the usability of cost-effective monitors for athletics events’ organisers, to minimise exposure of participants and spectators, avoid impacts on athletic performance and facilitate decision-making regarding the potential cancellation, postponement or relocation of competitions due to air quality issues. Currently, guidelines to minimise air pollution exposures during sports events are mostly non-existent. To this end, we implemented a pilot study which deployed air quality monitors in the main athletics stadium in 6 major cities around the globe, for a 1-year period. Cost-effective air quality monitors (KunakAir, Kunak Technologies) making use of low-cost sensors for NO, NO2, O3, PMx and CO were deployed in the athletics stadiums of major cities in Europe, N America, Asia, Australia, and Africa (2 cities). We conclude that the hyper-local data generated by the sensors were useful to describe daily air pollutant trends and identify hourly maxima for the different pollutants under study. This information has high added value for event organisers in terms of (1) identifying optimal times and seasons for competitions, (2) setting thresholds to decide on postponement or cancellation of events, and (3) application of targeted mitigation strategies. Absolute pollutant concentrations can only be compared directly across if sensors are calibrated locally, as sensor performance is impacted by local meteorology and air pollution mix (e.g., particle hygroscopicity, density, chemical composition).

Additional Authors: Mar Viana, Kostas Karatzas, Thanos Arvanitis, Cristina Reche, Miguel Escribano, Edurne Ibarrola, Paolo Adami, Stéphane Bermon

In-Person Poster Display: #41

The Potential of Air Sensors for Personalizing and Advancing Human Health Research

Development and evaluation of a Low cost TVOC sensor system for indoor and workplace exposure assessment.

Presented By: Alan Rossner, Clarkson University, The Institute for a Sustainable Environment

While VOCs are a broad array of compounds, health concerns with chronic exposure to compounds such as aromatics (BTEX) and chlorinated hydrocarbons are an ongoing concern in many indoor air environments. Indoor air concentrations of pollutants are often higher for TVOCs, CO and CO2 than outdoor concentrations. While established methods exist to collect time weighted average samples, future exposure assessment strategies should employ monitoring of VOCs with higher spatial and temporal resolution, even though low cost sensors will likely have a reduced data quality. The aim of this study is to evaluate three low cost TVOC sensors (Adafruit SGP 30, AMS 811 and AMS-MLV –P2) along with CO and CO2 sensors in chamber studies and field studies under a series of environmental conditions and concentrations. The chamber study design included exposing the sensors to 5 different VOCs at three different humidity levels (30, 50, and 80% RH), then three different temperatures (15, 22 and 30 oC). The performance of the low cost sensors was compared to both more expensive sensors and a GC/MS. To date, the results vary depending upon the type of sensor and type of VOC and to a lesser extent the environmental conditions of the chamber. In low relative humidity environments (<20%) there was little to no sensor response at known concentrations of each VOCs. In addition, responsiveness of the sensors has been observed to be chemical specific, with higher output signal in some VOCs (e.g. Toluene vs Trichloroethylene), when tested in known concentrations. Inter sensor variability was extensive, typically varying 6-7 times for each VOC. The chamber studies allow for a validation of performance, but also as a calibration of the different sensors. While laboratory testing is providing the performance in a controlled environment, field testing (in progress) will provide an assessment of the performance of each sampler under conditions similar to actual use. *Selected for a Lighting Talk

Additional Authors: Alan Rossner, Suresh Dhaniyala

In-Person Poster Display: #42

Mobile Sensing: A Quick-Start Guide to Equipping Vehicles with Air Quality Sensors

Presented By: Berj Der Boghossian, South Coast Air Quality Management District

Air quality sensors are becoming pivotal for developing new stationary networks while also expanding existing networks and advancing regional spatiotemporal mapping of air quality. In addition to fixed governmental air monitoring station networks, these stationary networks allow for large areas of unmonitored air quality to exist. Mobile monitoring is becoming an increasingly useful approach to collect useful air quality information in these unmonitored areas. California State legislation (i.e., Assembly Bill 617) requires air quality monitoring in communities heavily impacted by nearby air pollution sources. The Air Quality Sensor Performance Evaluation Center (AQ-SPEC) at the South Coast Air Quality Management District has been developing mobile platform vehicles equipped with regulatory-/research-grade analyzers and consumer-grade sensors. The AQ-SPEC scope of the mobile platform development was centralized on creating a Mobile Sensor Performance Evaluation Protocol from which new and innovative methods of equipping vehicles with sensors emerged. A standardized “quick-start guide” to equipping a vehicle with sensors to properly sample and measure air pollutants while on a mobile platform was created to aid community scientists to conduct air monitoring studies at the neighborhood level to better understand daily personal exposure to air pollution on their own.

Additional Authors: Berj Der Boghossian, Wilton Mui, Ashley Collier-Oxandale, Vasileios Papapostolou, Andrea Polidori

In-Person Poster Display: #43

Zooming in to zoom out: What we can learn about health risks from hyper local measurements

Presented By: Kristy Richardson, Colorado Department of Public Health and Environment

Understanding the impacts of air pollutants on human health remains a challenge of both spatial and temporal scales. While air regulations have historically focused on regional scales, we increasingly recognize the importance of zooming in to finer scales to meet people where they live and breathe. At the Colorado Department of Public Health and Environment, we use multiple technologies to get data representative of community exposures. This includes a range of spatial scales, from low cost photoionization detectors deployed in community members’ yards to high-tech mobile vans with remote-sensing equipment. Each type of data brings unique opportunities and challenges. We will summarize our approach to assessing health risks based on these different data sources. Similar challenges exist in traditional approaches to understanding health risks, which often had a zoomed out view, considering more serious outcomes and long-term exposures. Based on what we hear from people living with concerns about air pollutants, we know we need to zoom in to better understand short-term health impacts from very short exposures. These health impacts, like headaches and nosebleeds, are not often well represented in toxicity studies. We will discuss our approach to collecting meaningful data given these constraints and identify opportunities for future improvements. Finally we’ll reflect on how to zoom out, while incorporating the lessons learned at the more local scale to create better policy and health outcomes. *Selected for a Lighting Talk

Additional Authors: Kristy Richardson, Tara Webster

In-Person Poster Display: #46

A Paradigm Shift of Air Quality Monitoring, Requirements for the Next Generation Optical Particle Sensor and its Applications

Presented By: Raj Seelam, Piera Systems Inc.

For decades, air quality measurements have been synonymous with measuring PM2.5 and PM10, accumulated total weight of particles smaller than 2.5 and 10 microns in size. While PM2.5 and PM10 provide an overview of air quality, the metric of mass concentration in weight per volume (typically ug/m3) paints an incomplete picture of what is in the air. This metric has an inherent limitation of representing harmful particles which are dominant in number while their contribution to weight is insignificant, such as submicron particles. People spend majority of their time indoors these days and it’s dominated by particles smaller than PM2.5. Decades of research demonstrates that these smaller particles are significantly more dangerous to human health, thus necessitating better measurement and mitigation of these harmful particles in real-time. However, affordable low-cost sensors underperform in terms of accurate particle size measurement, while accurate devices are not widely accessible to the public due to high cost. This paper proposes a shift in paradigm of the next generation optical air quality measurement to include accurate count and size for various sized particles, and provides an overview of how precise and independent data with high resolution and granularity leads to possible identification of sources, and assess steps taken to mitigate the health impact.

Additional Authors: Aaron Soh, Vincent Ratford, Raj Seelam

In-Person Poster Display: #48

Comparison of Low-Cost Pollution Sensors Against Industrial Mining Dusts in a Calm-Air Aerosol Chamber

Presented By: Justin Patts, CDC / NIOSH

Low-cost dust monitors (LCDM), developed and marketed for air pollution monitoring are being implemented in communities to help understand the local and even hyper-local air quality. While prices vary, most of these integrated sensing packages cost less than $300 USD. In mining environments, area and wearable dust monitors have been around for decades but still average around $5000 per unit and are cost-prohibitive to deploy in sufficient quantity to enable a rich, real time understanding of mine dust levels. The objective of this study is to determine the potential suitability of low-cost dust sensors for applications within the mining industry. To accomplish this, calm-air aerosol chamber testing was conducted comparing the time-integrated average output of low-cost dust sensors to reference measurements while exposed to dust types commonly found in the metal/non-metal mining sector. Eight low-cost dust monitors and one industrial wearable dust monitor were simultaneously exposed to five different types of dust each on a different day. To quantify the performance of each sensor versus the reference we performed a first order least squares regression and report both the R2 as well as the correction factor (slope). The mean R2 of the LCDM responses relative to the reference standard was 87.6 and 74.0 for the AM520 (statistically significant at p <0.005). The mean correction factor was 4.04 for all the LCDM and 1.9 for the AM520 across all dust types (p < 0.001). The data collected from the low-cost dust sensors show that they are capable under these conditions of predicting the concentrations of dust in a dynamic range, and future studies should further explore their capability to operate in real-world mining scenarios.

Additional Authors: Justin Patts, Kyle Louk, Emanuele Cauda, Cody Wolfe

In-Person Poster Display: #44

High-resolution air quality data for exposure management

Presented By: Karim Tarraf, Hawa Dawa GmbH

Given the increasing urgency of air pollution-related diseases and, on the other hand, the advances made in air quality measurements and data science, this paper introduces a study on the use of air quality management for an exposure reduction of risk groups. The study was done for the Belfast Department of Health as part of a SBRI project aimed at conceptualising and developing digital solutions that utilise high-resolution air pollution data to guide urban planning, health service developments, and self-management of citizens and patients with risk factors. On the city level, the focus was on prioritising measures to minimise risk, separating unavoidable pollution and people at risk, and on mobility management. On the individual level, the focus was on avoiding routes with high pollution and planning activities at low-exposure times and locations. The core of the study was the correlation of high-resolution air quality data with demographic data, health (prescription) data and various points of interest categories. Based on these analytics, dashboards for urban planning and city-level health management were set up as well as a mobile application for risk group self-management. The study showed that high-resolution air pollution data combined with other data sources such as population and health data allow us to understand where people with different risk profiles are exposed to air pollution. This information can provide a basis for quantifying the benefits of different interventions regarding the health risks of air pollution. Various tools utilising high-resolution air pollution data can support exposure management on an individual or city level. An air quality management approach focused on exposure reduction of those at risk can provide a basis for effectively prioritising different, potentially dynamic intervention choices.

Additional Authors: Karim Tarraf, Birgit Fullerton

In-Person Poster Display: #45

Using air quality sensors to better quantify — and fight back against — the negative health impacts of exposure to wildfire smoke

Presented By: Ryan Higgins, Clarity

A growing body of scientific research is finding that the air pollution resulting from wildfire smoke carries deadly health impacts far beyond the heat and flames of the fires themselves. These public health impacts represent one of the greatest societal costs of destructive, uncontrolled wildfire — but to date, these costs have not been effectively measured or communicated to decision-makers or communities, largely due to a lack of sufficiently granular air quality data. Blue Forest and partners are working to better quantify the air quality-related public health costs of wildfire smoke and more explicitly connect forest restoration work that reduces the risk of severe wildfire with these costs. To support this effort, Blue Forest and the California Council on Science and Technology (CCST) were recently awarded an Innovative Finance for National Forests Grant (IFNF) to develop a cost-benefit model of reduced wildfire smoke impacts with forest management. As part of the IFNF award and improved air quality monitoring, Blue Forest has installed air quality sensors at sites in fire-prone areas across Northern California. The sensors are part of a broader network being deployed by the Feather River Air Quality Management District (FRAMQD), which will allow the District to better document how smoke is impacting communities in the area — as well as to be more proactive in providing alerts and advisories to the public when needed. Presenters will discuss how air quality sensors are being used to study the health impacts of wildfire smoke at a more granular level than previously possible — and to develop a cost-benefit model of reduced wildfire smoke impacts with forest management.

Additional Authors: Dr. Kim Seipp, Dr. Teresa Feo, Sean Wihera

In-Person Poster Display: #49

Youth-Focused Education and Youth-Lead Initiatives

Making Air Pollution Visual – Educational Resources using Air Sensors to Explore Air Quality

Presented By: Andrea Clements, U.S.EPA

Poor air quality is not always visible to the human eye. Since it is hard to see, many people are unaware of the risks associated with their daily exposures. Air sensors can provide visual information about air quality by displaying colors and/or numbers that correspond to pollutant concentrations, making them powerful tools for educational efforts. To support educators and air sensor loan pilot programs, U.S. EPA created five lesson plans designed to introduce air quality concepts including (1) outdoor air quality, (2) indoor air quality, (3) personal exposure, (4) vegetative barriers, and (5) wildfire smoke. The first 4 lessons can be conducted with any handheld sensor that measures and/or interactively maps fine particulate matter (PM2.5), allowing participants to explore how pollution levels change in their community. The last lesson uses crowdsourced PM2.5 sensor data displayed on a map alongside satellite fire detections and smoke observations to allow participants to investigate the impacts of wildfire smoke in their community. Each lesson plan includes background information, hands-on activities to explore air quality conditions and pollution sources and provides questions that help participants consider how air quality can be improved. Each lesson plan includes an instructor guide, participant guide, introductory slides, worksheet, resource list, frequently asked question document, and an alignment to the Next Generation Science Standards (NGSS). Disclaimer: Although this abstract was reviewed by EPA and approved for publication, it may not necessarily reflect official Agency policy. Mention of trade names or commercial products does not constitute endorsement or recommendation for use. *Selected for a Lighting Talk

Additional Authors: Andrea Clements, Rachelle Duvall, Olivia Ryder, Hilary Minor, Steve Brown

In-Person Poster Display: #52

RELAQS: Research and Education with Low-cost Air Quality Sensors

Presented By: Ben Crawford, University of Colorado Denver

Low-cost air quality sensors have tremendous potential as tools to support learning, in addition to their use for research and monitoring applications. Although low-cost sensors are increasingly used in a variety of education-focused activities, there are few studies designed to specifically assess the efficacy of low-cost sensors to help students meet course learning outcomes. In this project, custom built low-cost sensor kits are evaluated in a high school Advanced Placement (AP) environmental science course during the spring 2022 semester. The sensor kits measure particulate matter, VOCs, and carbon dioxide and each unit is powered by solar panels and rechargeable batteries. Sensor measurements are transmitted to an online database where data are visualized and downloadable in near real-time using a custom interactive dashboard interface. For this research, new classroom activities incorporating the sensors and their data are designed to meet specific US NGSS (Next Generation Science Standards) and AP standards. The evaluation study (two classes, ~25 students each) is designed so that one group of students participates in the sensor-based activities while a control group engages in traditional non-sensor activities. Progress towards learning outcomes in both groups is assessed by administering concept inventory quizzes before and after the activities, through responses to a survey instrument, and through student interviews. Here we present preliminary results from this study and ideas for future research directions, towards the overall goal to improve K-12 and undergraduate STEM education. *Selected for a Lighting Talk

Additional Authors: Ben Crawford, Bud Talbot, Kirsten Christensen, Shawndra Fordham, Eliza Rayner, Dan Connors

In-Person Poster Display: #57

Making the Invisible Visible: Blending STEM and Air Quality for Student Learning

Presented By: Olivia Ryder, Sonoma Technology

Improved air quality sensor technology is creating new educational opportunities. The Kids Making Sense® (KMS) air quality curriculum unites STEM education with hands-on projects and mobile air sensors to teach students about air pollution and how to reduce exposure. The KMS curriculum meets Next Generation Science Standards and Common Core Standards for grades 6-12 and has been successfully implemented in over 300 classrooms worldwide. Students learn about air pollution, particulate matter (PM), sources of PM, the health effects of PM, and regulatory monitoring. Working in teams, students design a study to monitor PM around their communities, use air sensors to collect credible air quality data, and analyze their findings using interactive online mapping tools. Our new AQ-go PM sensors allow students to measure air pollution, and are transparent to allow them to examine the sensor components. Finally, students develop an air quality awareness campaign to share what they’ve learned with members of their school or community. In parallel with the KMS air quality measurement program, and in partnership with the Blue Lake Rancheria Tribe, we developed a Build A Sensor Kit and lesson module that allows students to dive deeper into learning how particle sensors work. Participants learn about the components that comprise an air quality sensor, the electronics and data transfer process, and how to program their own sensors. Students learn valuable skills in engineering, electronics, and coding in this hands-on module. Providing students with the components and guidance to build their own sensor empowers them to be curious and spurs student-initiated inquiry. In this talk, we will describe the programs above, discuss key outcomes from teachers who have implemented the programs, and highlight the strengths of using hand-held sensor technology in classrooms or as outreach to engage students and empower them to create change. *Selected for a Lighting Talk

Additional Authors: Olivia Ryder, Josette Marrero, Tami Lavezzo, Will McCoy

In-Person Poster Display: #59

Algorithmic Correction of MOS Gas Sensor for Ambient Temperature and Relative Humidity Fluctuations

Presented By: Akarsh Aurora, Ashland High School

Algorithmic Correction of MOS Gas Sensor for Ambient Temperature and Relative Humidity Fluctuations

Additional Authors: Akarsh Aurora

In-Person Poster Display: #53

Lessons Learned From a Clean Air Equity Pilot for Students in Low Income Communities

Presented By: Andrew Clark, Sustainable Silicon Valley

Recent events surrounding the COVD pandemic have underscored the fact that low-income communities are disproportionately exposed to poor air quality. The Environmental Protection Agency and CARB have identified in Disadvantaged (SB535) and Low Income Communities (AB1550), (https://ww3.arb.ca.gov/cc/capandtrade/auctionproceeds/communityinvestments.htm [https://ww3.arb.ca.gov/cc/capandtrade/auctionproceeds/communityinvestments.htm] ). In response, Sustainable Silicon Valley has created the Clean Air Equality Project (CAEP) to conduct a community air monitoring program focused on middle and high school students in disadvantaged and low-income communities. Using low-cost sensors to gather hyper local data, we engaged students from diverse communities across the SF Bay Area who live with divergent air quality. CAEP participants use lightweightportable, smartphone-linked air quality sensor technologies to perform measurements, as local federal reference measurements are sparse. (https://www.backpaqlabs.com/docs/BackpAQ Personal Air Quality Monitor User Guide V0.98.1.pdf [https://www.backpaqlabs.com/docs/BackpAQ%20Personal%20Air%20Quality%20Monitor%20User%20Guide%20V0.98.1.pdf]) Low-cost sensors have been making great advances and offer the potential to obtain hyper-local information on air quality. Incorporating them into a user-friendly, easy-to-carry design is challenging, as is ensuring adequate data and spatial privacy. While these low-cost sensors do not have the accuracy and precision of federal reference methods, with proper inter-calibration and interpretation, they can shed light on the different air quality being experienced by diverse socio-economic communities.

Additional Authors: Andrew Clark, Anthony Strawa, Tim Dye

In-Person Poster Display: #55

A Maker-Friendly Design for Mobile Participatory Air Quality Monitoring

Presented By: Andrew Clark, Sustainable Silicon Valley

The availability of low-cost air quality sensors has made it possible to reimagine the role citizens, and in this case high school students, can play in monitoring and understanding their exposure to localized air pollution in their communities. To enable these students to engage in environmental justice projects and activities, new designs are emerging for portable, personal monitors that are easy-to-use, can be carried in hand, slung over bicycle handlebars, or clipped to a backpack. As these devices are typically targeted at non-professionals, ease-of-use and user-experience (UX) are paramount and must be baked into the design from day one. They must be rugged and fault-tolerant, as operating failures will quickly wear down the patience and good will of even the most eager participants. And to more deeply engage STEM-oriented makers, easy-to-build kits and accompanying instructions must be available to those who choose to construct their own instruments. With these criteria in mind, and with an eye toward keeping cost/student low, our organization has designed and is piloting an exemplary device - BackpAQ - which is tethered wirelessly to an app running on the student’s ubiquitous smartphone. Through the app, users can manage all functions, view real-time sensor values through gauges, graphs and charts, and interact with a neighborhood map showing PM2.5 or CO2 levels at key measurement points along their journey. Data is securely uploaded to a cloud-based database at regular intervals. To enhance students' understanding of data they have collected, a powerful open source-based air quality portal - BP AQView -provides a web-based analytics and data visualization toolset. There is provision as well for portable data interchange with advanced statistics and data science tools such as MatLab and R. By architecting this system as a set of modular components, it is both extensible and scalable. Experience and results from recent pilots will be presented.

Additional Authors: Andrew Clark, Anthony Strawa

In-Person Poster Display: #56

Incorporating Personal Monitoring Utilizing Low-cost Sensors into the Undergraduate Curriculum

Presented By: Joshua Stratton, Rider University

Environmentally aware and motivated students are abundant in today’s undergraduate classrooms. While some of these talented young scientists may become professionals in environmental related fields, many of them could serve as leaders for community-led projects in the future. This study aimed to challenge undergraduate students to design, execute, and interpret a personal monitoring project using mobile particulate matter (PM) sensors. This project included experimental design, data collection, data harmonization, manipulation, analysis via a programming language, data visualization, and a final in-class dissemination to their peers. Students were provided with a battery powered Arduino microcontroller, Plantower PMS5003 PM sensor, relative humidity (RH) and temperature (T) sensor (DHT-22), and local data storage to conduct this work. Students used their mobile phone GPS to track their location for comparison with PM concentrations. Students operated these sensors and extracted data from local storage. Although several students reported no previous experience with programing languages, all students were able to organize, merge, analyze, and visualize data using simple R templates provided within the course and tools such as EPA’s Real Time Geospatial Data Viewer (RETIGO). Based on student surveys, students reported a positive experience and were likely to engage in future community-led projects. However, data management and manipulation were the most difficult task based on student feedback. While student feedback recommended help from a technical expert, this study suggests motivated individuals may be capable of minor data management.

Additional Authors: Joshua Stratton, Jessica Barone, Nicolas Kaltenhauser, Carissa Moore, Karoline Barkjohn, Andrea Clements, Samuel Frederick, Luis Lim, Olga Boyko, Jessica Munyan, Rachael Leta-Graham

In-Person Poster Display: #60

A Simple Low-Cost Air Sensor Package for classroom monitoring of air quality

Presented By: Ajith Kaduwela, CARB

We have designed and built a simple low-cost air sensor package based on an ESP-32 microcontroller. This package measures CO2. particulates, noise, and light levels, and total volatile organic compounds (tVOC). in addition to temperature, relative humidity, and barometric pressure. It also has an LCD display that shows pollutant concentrations. Our goal is to make this low-cost sensor package compliant with Assembly Bill 841 in California.

This low-cost sensor package will send data to a remote server for display. The real time data will be available to all students, faculty, administrators, and visitors to the school. More aggregated data will be disseminated to the public due to privacy issues.

In the absence of heavy pollution events such as wildfires, the CO2 concentrations, together with particulate measurements, can be used as a surrogate for the extent of air circulation. That intern would be a measure of HVAC functionality and the COVID safety in the school. Teachers and school administrators will also be able to monitor pollutant levels using this system during heavy pollution events such as nearby wildfires. This would provide pertinent information to the management of air quality in a school during a heavy pollution event.

Additional Authors: David Zhan, Rashmit Shrestha, michael Zhou, Sean Morris, Kevin Lu, Ajith Kaduwela

In-Person Poster Display: #54

Communication Strategies for Understanding, Insight, and Action

The Enhanced U.S. EPA Air Sensor Guidebook

Presented By: Andrea Clements, U.S.EPA

The knowledge base surrounding air sensor applications and use has rapidly developed over the past decade. Around 2012, the availability and use of air sensors began to expand rapidly. At this time, air sensor users had a wealth of questions including how to use sensors, select sensors for specific applications, understand sensor performance, collect data, interpret the data, and more. In response, the U.S. EPA developed the Air Sensor Guidebook in 2014. The Guidebook was designed to provide basic information on air quality, guidance on selecting appropriate sensors for a given application, considerations for data quality, and sensor performance needs for different applications. The Guidebook has been one of the most popular resources available on the U.S. EPA’s Air Sensor Toolbox website. Since the Guidebook was released 7 years ago, the use and understanding of sensors has continued to grow and more information regarding best practices, sensor use, sensor performance, and other key topics is now available. The U.S. EPA is updating the original Air Sensor Guidebook to reflect this new information and to continue to support users and manufacturers of sensors. This presentation will provide a highlight of the updates to the Guidebook including the addition of new topics such as sensor performance guidance, determining the purpose for monitoring, and planning and conducting an air quality monitoring study. Disclaimer: Although this abstract was reviewed by EPA and approved for publication, it may not necessarily reflect official Agency policy.

Additional Authors: Andrea Clements, Rachelle Duvall, Danny Greene, Tim Dye

In-Person Poster Display: #52