Seasonality and Source Apportionment of Nonmethane Volatile Organic Compounds at Boulder Reservoir, Colorado, Between 2017 and 2019

Abstract

Multiyear observations of 13 nonmethane volatile organic compounds (NMVOCs) were collected at the Boulder Reservoir in the Colorado Northern Front Range Metropolitan Area (NFRMA). We separate abundances of NMVOCs in the 2017–2019 data into source contributions using two approaches that have been applied to prior NFRMA datasets. Positive matrix factorization analysis identifies five NMVOC factors in winter, spring, and fall that correspond to long‐lived and short‐lived NMVOCs from regional oil and natural gas (O&NG) production, traffic, local shorter‐lived alkenes, and regional anthropogenic background. In summer, there is an additional biogenic NMVOC factor dominated by isoprene. The PMF model indicates that 79 ± 1% of C2‐C5 alkanes in winter and 84 ± 20% in summer are attributed to O&NG activities. Ethyne is largely from traffic with contributions ranging from 45 ± 6% in winter to 87 ± 32% in summer. Ethene and propene are associated with a potentially separate source of shorter‐lived alkenes that we cannot identify. The largest contributing sectors to the observed hazardous air pollutants (HAPs) differ substantially by species and season. Benzene is attributed to O&NG production, traffic and other industrial activities. Toluene is predominantly attributed to regional anthropogenic activities in all seasons. Of the HAPs quantified in this dataset, hexane stands out as largely attributed to O&NG production. Consistent with prior analyses, this work shows that the NFRMA is more strongly influenced by O&NG sources than many other U.S. urban regions.

Weekend‐Weekday Implications and the Impact of Wildfire Smoke on Ozone and Its Precursors at Boulder Reservoir, Colorado Between 2017 and 2019

“…The number of smoke-impacted samples identified using these criteria is similar to that determined for the full data collection period at Boulder Reservoir (January 1, 2017 to December 31, 2019) described in Pollack et al (2021). Briefly, using these criteria, we find that less than 10% of the hourly samples (N = 864) were smoke impacted.…”

“…LLC, 2020). Details of the Boulder Reservoir measurement site, data collection methods, and the NMVOC measurements are described in Pollack et al (2021). Details of the NMVOC measurement technique are described in Tanner et al (2006).…”

“…Compounds quantified in each sample include ethane, ethene, propane, propene, i-butane, acetylene, n-butane, i-pentane, n-pentane, hexane, isoprene, benzene, and toluene. The method uncertainty and method limit of detection (MDL) for the NMVOC measurements are presented in Pollack et al (2021).…”

“…Briefly, using these criteria, we find that less than 10% of the hourly samples (N = 864) were smoke impacted. A subtle difference in the number of smoke-impacted NMVOC samples (455 smoke-impacted samples used here compared to 466 samples used in Pollack et al, 2021) is the result of our choice to conservatively remove a few data points that straddle the smoked-free and smoke-impacted periods.…”

“…We also note that a majority of the NMVOC measurements collected on July 11, 2018 (a smoke-free day) were below the MDL. We include this day in our analysis of the NMVOC data here and in Pollack et al (2021), although we do not utilize it in our interpretation of days with MDA8 O 3 > 70 ppbv.…”