Tryner, JessicaL'Orange, ChristianMehaffy, JohnMiller-Lionberg, DanielHofstetter, Josephine C.Wilson, AnderVolckens, John2019-06-242019-06-242019https://hdl.handle.net/10217/195445http://dx.doi.org/10.25675/10217/195445This dataset consists of data collected during two laboratory evaluations of PurpleAir monitors and one field deployment of PurpleAir monitors co-located with portable filter samplers.The pre-deployment laboratory evaluation took place on 2018-08-20. The post-deployment laboratory evaluation took place on 2018-12-17. The goals of these evaluations were to: (a) assess whether the PurpleAir monitors responded linearly to NIST Urban Particulate Matter concentrations ranging from approximately 0 to 75 micrograms per cubic meter, (b) obtain laboratory-derived gravimetric correction factors for fine particulate matter (PM2.5) concentrations reported by PurpleAir monitors, (c) determine whether the response of the PurpleAir monitors to NIST Urban Particulate Matter changed over the duration of the field deployment, and (d) evaluate the precision of co-located PurpleAir monitors.The field deployment took place in Fort Collins, Colorado, USA between 2018-10-22 and 2018-12-06. The goals of the field deployment were to: (a) determine whether gravimetric correction factors derived from periodic co-locations with portable filter samplers (called "ASPEN boxes") improved the accuracy of 72-hour average PM2.5 concentrations reported by PurpleAir monitors (relative to conventional PM2.5 filter samplers operated at 16.7 L/min) and (b) compare 72-hour average PM2.5 concentrations measured using portable filter samplers and conventional filter samplers.The files associated with this dataset include: (1) the raw data recorded by the PurpleAir monitors during the two laboratory evaluations and the field deployment; (2) the raw data recorded by a tapered element oscillating microbalance (TEOM) during the two laboratory evaluations; (3) the raw data recorded by the ASPEN boxes during the field deployment; (4) a summary file describing the time-averaged concentrations reported by the PurpleAir monitors and the TEOM during the discrete concentration steps that comprised each laboratory evaluation; and (5) a summary file describing the average PM2.5 concentrations measured using the PurpleAir monitors, ASPEN boxes, and conventional filter samplers at each field site during each 72-hour sample period.Department of Mechanical EngineeringDepartment of StatisticsLow-cost aerosol monitors can provide more spatially- and temporally-resolved data on ambient fine particulate matter (PM2.5) concentrations than are typically available from regulatory monitoring networks; however, low-cost monitors—which do not measure PM2.5 mass directly and tend to be sensitive to variations in particle size and refractive index—sometimes produce inaccurate concentration estimates. We investigated laboratory- and field-based approaches for calibrating low-cost PurpleAir monitors against gravimetric filter samples. First, we investigated the linearity of the PurpleAir response to NIST Urban PM and derived a laboratory-based gravimetric correction factor. Then, we co-located PurpleAir monitors with portable filter samplers at 15 outdoor sites spanning a 3×3-km area in Fort Collins, CO, USA. We evaluated whether PM2.5 correction factors derived from periodic co-locations with portable filter samplers improved the accuracy of PurpleAir monitors (relative to reference filter samplers operated at 16.7 L/min). We also compared 72-hour average PM2.5 concentrations measured using portable and reference filter samplers. Both before and after field deployment, the coefficient of determination for a linear model relating NIST Urban PM concentrations measured by a tapered element oscillating microbalance and the PurpleAir monitors (PM2.5 ATM) was 0.99; however, an F-test identified a significant lack of fit between the model and the data. The laboratory-based correction factor did not translate to the field. Correction factors derived in the field from monthly, weekly, semi-weekly, and concurrent co-locations with portable filter samplers increased the fraction of 72-hour average PurpleAir PM2.5 concentrations that were within 20% of the reference concentrations from 15% (for uncorrected measurements) to 45%, 59%, 56%, and 70%, respectively. Furthermore, 72-hour average PM2.5 concentrations measured using portable and reference filter samplers agreed (bias ≤ 20% for 71% of samples). These results demonstrate that periodic co-location with portable filter samplers can improve the accuracy of 72-hour average PM2.5 concentrations reported by PurpleAir monitors.ZIPTXTPDFCSVengair qualityair pollutionparticulate matterPlantowerultrasonic personal aerosol samplerUPASDataset