Atmospheric histories and emissions of chlorofluorocarbons CFC-13 (CClF<sub>3</sub>), ΣCFC-114 (C<sub>2</sub>Cl<sub>2</sub>F<sub>4</sub>), and CFC-115 (C<sub>2</sub>ClF<sub>5</sub>)

Abstract

<p><strong>Abstract.</strong> Based on observations of the chlorofluorocarbons CFC-13 (chlorotrifluoromethane), ΣCFC-114 (combined measurement of both isomers of dichlorotetrafluoroethane), and CFC-115 (chloropentafluoroethane) in atmospheric and firn samples, we reconstruct records of their tropospheric histories spanning nearly 8 decades. These compounds were measured in polar firn air samples, in ambient air archived in canisters, and in situ at the AGAGE (Advanced Global Atmospheric Gases Experiment) network and affiliated sites. Global emissions to the atmosphere are derived from these observations using an inversion based on a 12-box atmospheric transport model. For CFC-13, we provide the first comprehensive global analysis. This compound increased monotonically from its first appearance in the atmosphere in the late 1950s to a mean global abundance of 3.18<span class="thinspace"></span>ppt (dry-air mole fraction in parts per trillion, pmol mol⁻¹) in 2016. Its growth rate has decreased since the mid-1980s but has remained at a surprisingly high mean level of 0.02<span class="thinspace"></span>ppt yr⁻¹ since 2000, resulting in a continuing growth of CFC-13 in the atmosphere. ΣCFC-114 increased from its appearance in the 1950s to a maximum of 16.6<span class="thinspace"></span>ppt in the early 2000s and has since slightly declined to 16.3<span class="thinspace"></span>ppt in 2016. CFC-115 increased monotonically from its first appearance in the 1960s and reached a global mean mole fraction of 8.49<span class="thinspace"></span>ppt in 2016. Growth rates of all three compounds over the past years are significantly larger than would be expected from zero emissions. Under the assumption of unchanging lifetimes and atmospheric transport patterns, we derive global emissions from our measurements, which have remained unexpectedly high in recent years: mean yearly emissions for the last decade (2007–2016) of CFC-13 are at 0.48 ± 0.15<span class="thinspace"></span>kt yr⁻¹ (&amp;gt; 15<span class="thinspace"></span>% of past peak emissions), of ΣCFC-114 at 1.90 ± 0.84<span class="thinspace"></span>kt yr⁻¹ (∼ 10<span class="thinspace"></span>% of peak emissions), and of CFC-115 at 0.80 ± 0.50<span class="thinspace"></span>kt yr⁻¹ (&amp;gt; 5<span class="thinspace"></span>% of peak emissions). Mean yearly emissions of CFC-115 for 2015–2016 are 1.14 ± 0.50<span class="thinspace"></span>kt yr⁻¹ and have doubled compared to the 2007–2010 minimum. We find CFC-13 emissions from aluminum smelters but if extrapolated to global emissions, they cannot account for the lingering global emissions determined from the atmospheric observations. We find impurities of CFC-115 in the refrigerant HFC-125 (CHF₂CF₃) but if extrapolated to global emissions, they can neither account for the lingering global CFC-115 emissions determined from the atmospheric observations nor for their recent increases. We also conduct regional inversions for the years 2012–2016 for the northeastern Asian area using observations from the Korean AGAGE site at Gosan and find significant emissions for ΣCFC-114 and CFC-115, suggesting that a large fraction of their global emissions currently occur in northeastern Asia and more specifically on the Chinese mainland.</p>