Modern halogenated inhalation anesthetics undergo little metabolization during clinical application and evaporate almost completely to the atmosphere. Based on their first measurements in a range of environments, from urban areas to the pristine Antarctic environment, we detect a rapid accumulation and ubiquitous presence of isoflurane, desflurane, and sevoflurane in the global atmosphere. Over the past decade, their abundances in the atmosphere have increased to global mean mole fractions in 2014 of 0.097 ppt, 0.30 ppt, and 0.13 ppt (parts per trillion, 10 −12 , in dry air), respectively. Emissions of these long-lived greenhouse gases inferred from the observations suggest a global combined release to the atmosphere of 3.1 ± 0.6 million t CO 2 equivalent in 2014 of which ≈80% stems from desflurane. We also report on halothane, a previously widely used anesthetic. Its global mean mole fraction has declined to 9.2 ppq (parts per quadrillion, 10 −15 ) by 2014. However, the inferred present usage is still 280 ± 120 t yr −1 .

Hydrochlorofluorocarbon HCFC‐133a (CF3CH2Cl) is an anthropogenic compound whose consumption for emissive use is restricted under the Montreal Protocol. A recent study showed rapidly increasing atmospheric abundances and emissions. We report that, following this rise, the atmospheric abundance and emissions have declined sharply in the past three years. We find a Northern Hemisphere HCFC‐133a increase from 0.13 ppt (dry‐air mole fraction in parts per trillion) in 2000 to 0.50 ppt in 2012–mid‐2013 followed by an abrupt drop to ∼0.44 ppt by early 2015. Global emissions derived from these observations peaked at 3.1 kt in 2011, followed by a rapid decline of ∼0.5 kt yr−2 to reach 1.5 kt yr−1 in 2014. Sporadic HCFC‐133a pollution events are detected in Europe from our high‐resolution HCFC‐133a records at three European stations, and in Asia from samples collected in Taiwan. European emissions are estimated to be <0.1 kt yr−1 although emission hot spots were identified in France.

We report the first multiyear atmospheric record of HCFC-31 (CH 2 ClF), based on flask samples and in situ analyses of air from both hemispheres. Although HCFC-31 has never been produced in large amounts, observed mole fractions in the Northern Hemisphere increased from 2000 onward, reaching 170 ppq (parts per quadrillion, 10 À15) in 2011-2012 before decreasing rapidly. By combining our observations with a two-dimensional atmospheric chemistry-transport model, we infer an increase in global emissions from 240 t yr À1 in 2000 to 840 t yr À1 in 2011-2012, followed by a relatively fast decline to 570 t yr À1 in 2014. Emissions of HCFC-31 originate most probably from intermediate product release during the manufacturing process of HFC-32 (CH 2 F 2). The rapid decline in recent years could be due to changes in production methods rather than declines in diffusive sources such as landfills or HCFC-31 contaminations in merchandised HFC-32.

Abstract. A wide range of anthropogenic halocarbons is released to the atmosphere, contributing to stratospheric ozone depletion and global warming. Using measurements of atmospheric abundances for the estimation of halocarbon emissions on the global and regional scale has become an important top-down tool for emission validation in the recent past, but many populated and developing areas of the world are only poorly covered by the existing atmospheric halocarbon measurement network. Here we present 6 months of continuous halocarbon observations from Finokalia on the island of Crete in the Eastern Mediterranean. The gases measured are the hydrofluorocarbons (HFCs), HFC-134a (CH 2 FCF 3 ), HFC-125 (CHF 2 CF 3 ), HFC-152a (CH 3 CHF 2 ) and HFC143a (CH 3 CF 3 ) and the hydrochlorofluorocarbons (HCFCs), HCFC-22 (CHClF 2 ) and HCFC-142b (CH 3 CClF 2 ). The Eastern Mediterranean is home to 250 million inhabitants, consisting of a number of developed and developing countries, for which different emission regulations exist under the Kyoto and Montreal protocols. Regional emissions of halocarbons were estimated with Lagrangian atmospheric transport simulations and a Bayesian inverse modeling system, using measurements at Finokalia in conjunction with those from Advanced Global Atmospheric Gases Experiment (AGAGE) sites at Mace Head (Ireland), Jungfraujoch (Switzerland) and Monte Cimone (Italy). Measured peak mole fractions at Finokalia showed generally smaller amplitudes for HFCs than at the European AGAGE sites except for periodic peaks of HFC-152a, indicating strong upwind sources. Higher peak mole fractions were observed for HCFCs, suggesting continued emissions from nearby developing regions such as Egypt and the Middle East. For 2013, the Eastern Mediterranean inverse emission estimates for the four analyzed HFCs and the two HCFCs were 13.9 (11.3-19.3) and 9.5 (6.8-15.1) Tg CO 2 eq yr −1 , respectively. These emissions contributed 16.8 % (13.6-23.3 %) and 53.2 % (38.1-84.2 %) to the total inversion domain, which covers the Eastern Mediterranean as well as central and western Europe. Greek bottom-up HFC emissions reported to the UNFCCC were higher than our top-down estimates, whereas for Turkey our estimates agreed with UNFCCC-reported values for HFC-125 and HFC-143a, but were much and slightly smaller for HFC-134a and HFC152a, respectively. Sensitivity estimates suggest an improvement of the a posteriori emission estimates, i.e., a reduction of the uncertainties by 40-80 % in the entire inversion domain, compared to an inversion using only the existing central European AGAGE observations.

15A wide range of anthropogenic halocarbons is released to the atmosphere, contributing to stratospheric ozone depletion and global warming. Using measurements of atmospheric abundances for the estimation of halocarbon emissions on the global and regional scale has become an important top-down tool for emission validation in the recent past, but many populated and developing areas of the world are only poorly covered by the existing atmospheric halocarbon and 9.7 (4.3-15.7) Tg CO2eq yr -1 , respectively. These emissions contributed 17.3% (7.9-27.4%) and 53% (23.5-86%) to the total inversion domain, which covers the Eastern Mediterranean as well as Central and Western Europe. Greek bottom-up HFC emissions reported to the UNFCCC were much smaller than our top-down estimates, whereas for Turkey our estimates agreed with 40 UNFCCC-reported values for HFC-125 and HFC-143a, but were much and slightly smaller for HFC-134a and HFC-152a, respectively. Sensitivity estimates suggest an improvement of the a posteriori emission estimates, i.e. a reduction of the uncertainties by 40-80%, compared to an inversion using only the existing Central European AGAGE observations. Atmos. Chem. Phys. Discuss., https://doi