The Montreal Protocol was designed to protect the stratospheric ozone layer by enabling reductions in the abundance of ozone-depleting substances such as chlorofluorocarbons (CFCs) in the atmosphere1,2,3. The reduction in the atmospheric concentration of trichlorofluoromethane (CFC-11) has made the second-largest contribution to the decline in the total atmospheric concentration of ozone-depleting chlorine since the 1990s1. However, CFC-11 still contributes one-quarter of all chlorine reaching the stratosphere, and a timely recovery of the stratospheric ozone layer depends on a sustained decline in CFC-11 concentrations1. Here we show that the rate of decline of atmospheric CFC-11 concentrations observed at remote measurement sites was constant from 2002 to 2012, and then slowed by about 50 per cent after 2012. The observed slowdown in the decline of CFC-11 concentration was concurrent with a 50 per cent increase in the mean concentration difference observed between the Northern and Southern Hemispheres, and also with the emergence of strong correlations at the Mauna Loa Observatory between concentrations of CFC-11 and other chemicals associated with anthropogenic emissions. A simple model analysis of our findings suggests an increase in CFC-11 emissions of 13 ± 5 gigagrams per year (25 ± 13 per cent) since 2012, despite reported production being close to zero4 since 2006. Our three-dimensional model simulations confirm the increase in CFC-11 emissions, but indicate that this increase may have been as much as 50 per cent smaller as a result of changes in stratospheric processes or dynamics. The increase in emission of CFC-11 appears unrelated to past production; this suggests unreported new production, which is inconsistent with the Montreal Protocol agreement to phase out global CFC production by 2010.
Subscribe to Journal
Get full journal access for 1 year
only $3.90 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
Carpenter, L. J. et al. in Scientific Assessment of Ozone Depletion: 2014. Global Ozone Research and Monitoring Project—Report No. 55, Ch. 1, 1.1–1.101, (World Meteorological Organization, Geneva, 2014).
Montzka, S. A. et al. Decline in the tropospheric abundance of halogen from halocarbons: Implications for stratospheric ozone depletion. Science 272, 1318–1322 (1996).
Prinn, R. G. et al. A history of chemically and radiatively important gases in air deduced from ALE/GAGE/AGAGE. J. Geophys. Res. 105, 17751–17792 (2000).
Ozone Secretariat. Handbook for the Montreal Protocol on Substances that Deplete the Ozone Layer, 9th edn (United Nations Environment Programme, 2012).
Technology and Economic Assessment Panel. Assessment of Alternatives to HCFCs and HFCs and Update of the TEAP 2005 Supplement Report Data. Task Force Decision XX/8 Report (United Nations Environment Programme, Nairobi, 2009).
Daniel, J. et al. in Scientific Assessment of Ozone Depletion: 2010. Global Ozone Research and Monitoring Project—Report No. 52, Ch. 5, 5.1–5.56 (World Meteorological Organization, Geneva, 2011).
Harris, N. R. P. in et al. Scientific Assessment of Ozone Depletion: 2014. Global Ozone Research and Monitoring Project—Report No. 55, Chapter 5, 5.1–5.58 (World Meteorological Organization, Geneva, 2014).
Montzka, S. A. et al. in Scientific Assessment of Ozone Depletion: 2010. Global Ozone Research and Monitoring Project—Report No. 52, Chapter 1, 1.1–1.108 (World Meteorological Organization, Geneva, 2011).
Liang, Q. et al. Constraining the carbon tetrachloride (CCl4) budget using its global trend and inter-hemispheric gradient. Geophys. Res. Lett. 41, 5307–5315 (2014).
Lin, M., Horowitz, L. W., Oltmans, S. J., Fiore, A. M. & Fan, S. Tropospheric ozone trends at Mauna Loa Observatory tied to decadal climate variability. Nat. Geosci. 7, 136–143 (2014).
Oram, D. E. et al. A growing threat to the ozone layer from short-lived anthropogenic chlorocarbons. Atmos. Chem. Phys. 17, 11929–11941 (2017).
Rosenlof, K. R. & Reid, G. C. Trends in the temperature and water vapor content of the tropical lower stratosphere: Sea surface connection. J. Geophys. Res. 113, D06107 (2008).
Randel, W. J., Wu, F., Vömel, H., Nedoluha, G. E. & Forster, P. Decreases in stratospheric water vapor after 2001: Links to changes in the tropical tropopause and the Brewer-Dobson circulation. J. Geophys. Res. 111, D12312 (2006).
Ploeger, F. et al. Variability of stratospheric mean age of air and of the local effects of residual circulation and eddy mixing. J. Geophys. Res. 120, 716–733 (2015).
Stiller, G. P. et al. Shift of subtropical transport barriers explains observed hemispheric asymmetry of decadal trends of age of air. Atmos. Chem. Phys. 17, 11177–11192 (2017).
Newman, P. A., Coy, L., Pawson, S. & Lait, L. R. The anomalous change in the QBO in 2015–2016. Geophys. Res. Lett. 43, 8791–8797 (2016).
Chirkov, M. et al. Global HCFC-22 measurements with MIPAS: retrieval, validation, global distribution and its evolution over 2005–2012. Atmos. Chem. Phys. 16, 3345–3368 (2016).
Marsh, D. R. et al. Climate change from 1850 to 2005 simulated in CESM1(WACCM). J. Clim. 26, 7372–7391 (2013).
Lamarque, J.-F. et al. CAM-chem: description and evaluation of interactive atmospheric chemistry in the Community Earth System Model. Geosci. Model Dev. 5, 369–411 (2012).
Rienecker, M. M. et al. MERRA: NASA’s modern-era retrospective analysis for research and applications. J. Clim. 24, 3624–3648 (2011).
Boothe, A. C. & Homeyer, C. R. Global large-scale stratosphere-troposphere exchange in modern reanalyses. Atmos. Chem. Phys. 17, 5537–5559 (2017).
Ashford, P. et al. in Safeguarding the Ozone Layer and the Global Climate System: Issues Related to Hydrofluorocarbons and Perfluorocarbons (eds Metz, B. et al.) Ch. 7 (Cambridge Univ. Press, Cambridge, 2005).
Hu, L. et al. Considerable contribution of the Montreal Protocol to declining greenhouse gas emissions from the United States. Geophys. Res. Lett. 44, 8075–8083 (2017).
Laube, J. C. et al. Newly detected ozone-depleting substances in the atmosphere. Nat. Geosci. 7, 266–269 (2014).
Hossaini, R. et al. The increasing threat to stratospheric ozone from dichloromethane. Nat. Commun. 8, 15962 (2017).
Montzka, S. A. et al. in Scientific Assessment of Ozone Depletion: 2002. Global Ozone Research and Monitoring Project—Report No. 47, Ch. 1, 1.1–1.83 (World Meteorological Organization, Geneva, 2003).
Jones, A. R., Thomson, D. J., Hort, M. & Devenish, B. in Air Pollution Modeling and its Application XVII (eds Borrego C. & Norman A.-L.) Ch. 62, 580–589 (Springer, New York, 2007).
Hall, B. D., Dutton, G. S. & Elkins, J. W. The NOAA nitrous oxide standard scale for atmospheric observations. J. Geophys. Res. 112, D09305 (2007).
McCulloch, A., Ashford, P. & Midgley, P. M. Historic emissions of fluorotrichloromethane (CFC-11) based on a market survey. Atmos. Environ. 35, 4387–4397 (2001).
Liang, Q., Stolarski, R. S., Douglass, A. R., Newman, P. A. & Nielsen, J. E. Evaluation of emissions and transport of CFCs using surface observations and their seasonal cycles and the GEOS CCM simulation with emissions-based forcing. J. Geophys. Res. 113, D14302 (2008).
Rigby, M. et al. Re-evaluation of the lifetimes of the major CFCs and CH3CCl3 using atmospheric trends. Atmos. Chem. Phys. 13, 2691–2702 (2013).
Hamilton, K. & Fan, S.-M. Effects of the stratospheric quasi-biennial oscillation on long-lived greenhouse gases in the troposphere. J. Geophys. Res. 105, 20581–20587 (2000).
Simmonds, P. G. et al. Interannual fluctuations in the seasonal cycle of nitrous oxide and chlorofluorocarbons due to the Brewer-Dobson circulation. J. Geophys. Res. 118, 10694–10706 (2013).
Nevison, C. D. et al. Exploring causes of interannual variability in the seasonal cycles of tropospheric nitrous oxide. Atmos. Chem. Phys. 11, 3713–3730 (2011).
We thank NOAA station personnel for sample flask collection and on-site instrument operation, maintenance and troubleshooting; and personnel from cooperative institutions involved with flask sampling in Australia (CSIRO), Canada (AES), Ireland (University of Bristol), Israel (Weizmann Institute) and the United States (University of Colorado, Harvard University, University of Wisconsin, and the Scripps Institute for Oceanography). We thank the US National Science Foundation for logistics support at Summit (Greenland) and the South Pole; J. Butler, D. Fahey, S. Reimann, P. Newman and scientists from the Advanced Global Atmospheric Gases Experiment for discussions; S. Davis for MERRA2 reanalysis winds; and P. Novelli for CO data from MLO. The CESM project is supported by the NSF and the Office of Science (BER) of the US Department of Energy. We acknowledge the NOAA Research and Development High Performance Computing Program for computing and storage resources. This work was funded in part by the NOAA Climate Program Office’s AC4 program. The scientific results and conclusions, as well as any views or opinions expressed herein, are those of the authors and do not necessarily reflect the views of NOAA or the Department of Commerce.
Nature thanks J. Laube and the other anonymous reviewer(s) for their contribution to the peer review of this work.
The authors declare no competing interests.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data figures and tables
Extended Data Fig. 1 Hemispheric differences in CFC-11 mole fractions represented by results from individual sites at comparable latitudes.
a, Twelve-month running means of monthly differences are plotted at the mid-point of that time interval. Results from low latitudes (green lines) include a site at high-altitude (MLO) and low-altitude (KUM) in the Northern Hemisphere compared to the site at American Samoa (SMO). Results from mid- to high-latitude site pairs are indicated in other colours and include data from high-altitude (NWR, SUM, SPO) and low-altitude (THD, MHD, BRW) sites. Comparisons made at sites with similar sampling altitudes are indicated in bold lines. b, Details of site locations from which measurements of CFC-11 are obtained from flasks and from in situ instrumentation.
Extended Data Fig. 2 Observed and simulated global rates of change and hemispheric differences for some other long-lived chemicals.
a, Measured global surface rates for N2O (grey line), CFC-12 (thin blue lines), and CFC-113 (thin green lines) from flasks analysed by GC–ECD and also, for the CFCs, by GC–MS. b, Hemispheric differences measured for CFC-12 and N2O. c, Hemispheric differences measured for CFC-113. Multiple CCM simulation results appear in a, b, and c for CFCs as thick dark lines and are updated only annually; they represent simulations using the CAMCHEM model with MERRA2 reanalysis meteorology and the three-box-derived emission history. Dashed lines after 2012 represent simulations with emissions kept constant after 2012 (dark blue for CFC-12 or dark green for CFC-113), or when the three-box-derived emission record was considered but dynamics in 2012 were repeated in subsequent years (red dashed lines). Emission distribution 1 was used in all simulations (see Methods). Interannual variability in global growth rates for these gases are sometimes correlated, suggesting a common cause related to STE dynamics, perhaps associated with the quasi-biannual oscillation, although emission variations are particularly likely for N2O (for example, refs 32,33,34). This may explain the peak in growth rates for a number of gases in 2015. The change in rate for CFC-11 (see Fig. 4), however, is substantially larger and is sustained in 2016 when rates for other gases do not change appreciably or become smaller, suggesting that the underlying causes for the majority of changes observed for CFC-11 are unique to that gas.
Extended Data Fig. 3 The sensitivity of hemispheric mole fraction differences to variations in hemispheric air mass exchange.
a, b, Points represent the observation-based (blue symbols) or modelled (red and grey symbols) hemispheric difference as a function of the global emission rate derived for that year in the three-box model for HCFC-22 (a) and for HFC-134a (b; see Methods; lines connect sequential years and the legend applies to a and b). The sensitivity of the hemispheric mole fraction difference (N−S difference) to exchange timescale (τexch, N−S) was tested in the model by incorporating values of ± 0.1 yr around 1.1 yr. If this timescale did not vary interannually, we would expect the observation-based points (blue) to overlay those from the model (red). A change in the annual mean value of this exchange would increase the difference between the observed and modelled N−S difference. Specifically, an annual mean change of ± 0.1 year would be reflected in the observed N−S difference being two-thirds of the way closer to the grey point associated with the emission derived for that year. The consistency between the model (red) and observed (blue) hemispheric differences in most years suggests that interannual changes in the exchange timescale are 0.1 yr (around 10%) or less, typically. More importantly, the results show no systematic change in this relationship before and after 2012, suggesting that any change in the rate of hemispheric air exchange in the troposphere is less than 10% during this period. We estimate that to fit the observed increase in the N–S difference measured for CFC-11 after 2012 without increasing the net CFC-11 flux to the Northern Hemisphere, this exchange time constant would have had to increase from 1.1 to 1.7 yr, which is inconsistent with the results presented here. Although the distribution of emissions between and within the hemispheres can affect the N–S difference, any considerable change in this distribution over time would probably be a shift to lower latitudes (away from the US and Europe) and would lead to a decrease in the N−S difference over time, not an increase as is observed for CFC-11 after 2012. Consistent with this, the best fit to the observations was obtained when the emission distribution (North Hemisphere/global) in these analyses was linearly decreased over time (from 0.95 in 1995 to 0.85 in 2015 for HFC-134a, and from 0.86 to 0.82 for HCFC-22). Assuming a constant hemispheric emission distribution (Northern Hemisphere/global) over time does not change the conclusions from this analysis.
Extended Data Fig. 4 Measured and modelled annual hemispheric differences versus global emissions of CFC-11.
a, Measured mole fraction differences between the Northern and Southern Hemispheres (North − South) as a function of the global emission derived with a three-box model for 1978–2016; the line is a fit to all results and each point represents an annual mean difference and emission for a particular year. b, An expanded scale of data displayed in a with results from different measurement methods represented by symbols of the same colour; grey symbols (plus signs and diamonds) refer to a combined set of results from flasks and in situ instruments analysed by GC–ECD. For each method (colour), unfilled symbols refer to results for the years 2010–2012; filled symbols refer to 2013–2016. Specific years are labelled for GC–MS results during 2013–2016 and for ECD results during 1997–2000 (for example, ’15 is 2015). The data show that the relationship observed here during 2014–2016 is similar to that observed during 1996–2000. c, Same as b, but with differences between the Northern and Southern Hemispheres derived from the three-box model shown (black points and line connecting sequential years); select model years are labelled.
a, Measured mole fractions of CFC-11 and HCFC-22 in all samples collected during autumn (fraction of year > 0.6 and < 0.9) at MLO in the past nine years. b, Results for CH2Cl2 versus HCFC-22 in those same samples and years. c, The r2 regression coefficients (blue filled symbols, left-hand scale) and slopes (red unfilled symbols, right-hand scale) determined from the data in a over time. Only slopes for correlations that are significant at P < 0.05 are shown (that is, those for which r2 > ~0.25). d, The same as c, but for the data in b (CH2Cl2 versus HCFC-22). Eastern Asia has been a substantial source of HCFC-22 and CH2Cl2 for a number of years1,11. As a result, significant correlated variability is expected in their mixing ratios downwind of this region; this is borne out in observations at MLO during autumn from 2009 to 2017. These data may also provide rough estimates of relative emission magnitudes. For example, inventory- and atmosphere-based studies suggest emissions of HCFC-22 from China of around 100 Gg in 2010, increasing to 150 Gg in 2015 (ref. 1). Considering the slopes measured at MLO between HCFC-22 and CH2Cl2, this would correspond to regional emissions of CH2Cl2 of 300 Gg in 2010 increasing to 440 Gg in 2016. This is comparable to the 455 Gg ( ± 10%) estimated to have been used in China for emissive applications in 2015 (ref. 11). Applying the same analysis to CFC-11 suggests total emissions of 30–40 Gg yr−1 for 2014–2017, or 10–35 Gg higher than estimated for Chinese CFC-11 emissions in 2008–2009 (considering errors1), which is of the same order as the global CFC-11 emission increase derived here for 2014–2016 (13 ± 5 Gg yr−1). Although our data and analyses do not allow for a robust identification of the origin of the increase in CFC-11 emissions, we explore China’s potential contribution because it is also the largest producer and user of HCFCs in eastern Asia (see ref. 4 and http://ozone.unep.org/en/data-reporting/data-centre).
Extended Data Fig. 6 Correlations between additional trace gases measured during autumn at Mauna Loa.
Same as Extended Data Fig. 5, but for mole fractions of carbon monoxide versus HCFC-22 measured at MLO during autumn. a, The results in individual years. b, The r2 regression coefficients (blue filled symbols, left-hand axis) and slopes (red unfilled symbols, right-hand axis) determined from the data in a over the past eight years.
a, Mole fractions of HCFC-22 measured in flasks collected at MLO during the autumn of 2011 (red lines and symbols) and 2016 (blue lines and symbols). b, The same as a, but for CFC-11. c–f, Back trajectories calculated27 for 2011 samples indicated by the red text L1, L2, H1 and H2 in a and b. g–j, Back trajectories calculated for 2016 samples indicated by the blue text L1, L2, H1 and H2 in a and b. In c–j, darker shading represents surface regions sensed by the corresponding sampling events at MLO, with darker colours indicating greater influence. The colour scale in the trajectory maps is logarithmic (1 × 10−9 to 1 × 10−3 g s m−3, darker colours for higher concentrations) and represents the calculated time-averaged concentration within the 0–2,000 m surface layer during the 30 days before the sampling event given a point release at MLO27 of 1 g s−1. Increased mole fractions of HCFC-22 are observed in both 2011 and 2016 (labelled H1 and H2 in a or b and High 1 and High 2 in c–j) when surface sensitivity over eastern Asia is enhanced; CFC-11 mole fractions at MLO co-vary with HCFC-22 in these eastern-Asian influenced samples only after 2012. Some industrialized regions (for example, Japan) have considerable influence on samples containing both high and low mole fractions of CFC-11, HCFC-22 and CH2Cl2 and, therefore, are less likely to be the source of the greater mole fractions of CFC-11 at MLO after 2012. These results, along with results from Fig. 3 and Extended Data Figs. 5 and 6, suggest an increase in CFC-11 emissions from eastern Asia that is coincident with the increase in global emissions derived from our sampling network.
Extended Data Fig. 8 Additional model simulations of the changes in CFC-11 mole fraction over time and of hemispheric differences.
Rates of change and hemispheric differences from different combinations of emission distributions (E1, E2, E3), reanalysis meteorology (MERRA1 (M1), MERRA2 (M2), and GEOS5 (G5)), and CCMs (CAM and WACCM) are compared to quantities derived from observations (red lines or shading indicate the range of results from two (hemispheric differences) or three (global change rates) measurement techniques (Methods)). In all panels, results from observations and the CAM run using the Emission1 distribution and MERRA2 reanalysis meteorology are shown for reference (solid light blue and green lines). All blue lines represent simulations using the emission record derived from the three-box model analysis of observations, whereas all green lines indicate simulations with emissions kept constant after 2012 at the 2012 rate. a–c, Results from CAM as a function of emission distribution (E1 and E3) and nudging methodology (temperature and winds, or wind-only). d–f, Results from WACCM as a function of reanalysis meteorology (MERRA1 or MERRA2). g–i, Results from WACCM with GEOS5 reanalysis meteorology and two different emission distributions (E2 and E1). The comparisons are made for the CFC-11 global rate of change at the Earth’s surface (a, d, g; left column), the surface mean hemispheric difference (b, e, h; middle column), and the change in the surface mean hemispheric difference relative to the mean during 2010–2012 (c, f, i; right column; note expanded time axis). All quantities being compared are derived from hemispheric means determined from cosine of latitude weighting of observed or simulated mole fractions at sampling locations (Methods).
Extended Data Fig. 9 The sensitivity of derived bank release rates to CFC-11 lifetime and incineration.
Bank release rates derived with a 57.5-year lifetime as shown in Fig. 2b (black squares and the dashed blue line) are also calculated considering the upper and lower values for the most likely CFC-11 lifetime range (red lines; 43–67 yr, ref. 1). Including quantities of CFC-11 destroyed (for example, by incineration) reported to UNEP (grey circles) affect this result minimally.
About this article
Cite this article
Montzka, S.A., Dutton, G.S., Yu, P. et al. An unexpected and persistent increase in global emissions of ozone-depleting CFC-11. Nature 557, 413–417 (2018). https://doi.org/10.1038/s41586-018-0106-2
Nature Communications (2021)