Sources and Global Distribution of Atmospheric Perchlorate Aerosol

PEDRO CAMPUZANO-JOST, Dongwook Kim, Hongyu Guo, Benjamin A. Nault, Jason Schroder, Rainer Volkamer, Joseph Katich, Joshua P. Schwarz, Chelsea Thompson, Jeff Peischl, Thomas Ryerson, Andrew Weinheimer, Alessandro Franchin, Teresa Campos, Rebecca Hornbrook, Alan Hills, Eric Apel, Roisin Commane, Bruce Daube, Steven Wofsy, Glenn Diskin, Yuk Chun Chan, Lyatt Jaegle, Jose-Luis Jimenez, CIRES, University of Colorado, Boulder

     Abstract Number: 222
     Working Group: Aerosol Chemistry

Abstract
Perchlorate (ClO₄^-) is ubiquitous in the environment. Ice core data suggest it has been steadily increasing globally since the 1970s. The health implications of this increase are controversial. But importantly, we also know very little about the sources and chemistry driving this increase. There are some natural deposits of perchlorate, and perchlorate is produced industrially for rocket propellants. However, as recently reviewed by Chan et al. (2023), most of the available evidence suggests that neither of these sources can explain the increase. Rather, the deposited perchlorate is likely formed by atmospheric oxidative chemistry.

We have recently shown that specific, quantitative detection of perchlorate in ambient submicron aerosol is possible using an Aerodyne HR-AMS instrument. Subsequent analyses have refined our understanding of the instrument's response in polluted environments and led to significantly enhanced detection limits (0.25 pptv / 1 ng m^-3 @STP for 1 min sampling). This work leverages a reanalysis of a large set of NASA, NSF, and DOE campaigns to derive the first atmospheric dataset of perchlorate of near global scope up to altitudes of 14 km.

Perchlorate has been proposed previously as a byproduct of the stratospheric ClO_x cycle (Jaegle et al, 1996), which matches the strong stratospheric signature in our dataset (up to 3 pptv). GEOS-Chem simulations, with a simple modified ClO_x mechanism, are able to reproduce the stratospheric observations, but cannot account for the large tropospheric deposition.

Our dataset shows that significant amounts of perchlorate are detected downwind of large urban areas and open fires, with emission factors between 10-30 µmol/mol relative to CO, suggesting strong combustion sources of perchlorate. Implementing such sources in GEOS-Chem markedly improves overall model skill. Evidence for additional tropospheric sources of perchlorate, such as volcanic eruptions and boundary layer ozone depletion, and its implications will be discussed as well.