The group of halogenated greenhouse gases comprises chlorofluorocarbons (CFCs), hydrochlorofluorocarbons (HCFCs), hydrofluorocarbons (HFCs), chlorocarbons (e.g. CCl₄), bromocarbons (halons, H₃Br) and several other fluorinated gases (e.g. SF₆, NF₃). They are mainly used for applications such as refrigeration, air conditioning, foam blowing, and as solvents or propellants. Although they are satisfying the market’s specific requirements, all of these substances are contributing to climate change and the chlorine and bromine-containing substances additionally destroy ozone in the stratosphere. As a consequence both, the Montreal Protocol and the Kyoto Protocol are aiming at reducing the emissions of these compounds. The Montreal Protocol by totally banning emissive uses of ozone-depleting substances and recently by also including a reduction scheme for HFCs. The Kyoto Protocol by including fluorinated gases into the basket of compounds to be reported as emissions by each country on a yearly basis. Country emission reports are based on so-called bottom-up inventories that are produced by combining emissions factors for individual processes with statistical information on their activity. This approach may involve relatively large uncertainties due to uncertain or missing data on source activities and related emission factors. In addition, the bottom-up estimate inhibits the potential weaknesses of omitting important source processes altogether, leading to incompleteness and underreporting. Therefore in recent years, top-down approaches, which combine atmospheric measurements with atmospheric transport and chemistry models, have been used in support of bottom-up inventories.
As part of this study we performed two measurement campaigns in the Mediterranean, where until now no long-term measurements of halogenated compounds have been available. These campaigns provided the basis for the first top-down emission estimates of halocarbons of several industrialized and densely populated Mediterranean regions. The first campaign was performed in the Eastern Mediterranean at Finokalia (Crete) in 2012-2013. By combining measurements from Finokalia, Jungfraujoch, Mace Head and Mt. Cimone with meteorological transport models, emissions from cities and industrialized regions in Greece, Turkey and Egypt could be estimated. Emissions of the groups of analysed HFCs (HFC-134a, HFC-125, HFC-152a, HFC-143a) and HCFCs (HCFC-22, HCFC-142b) were estimated at 14.7 (6.7-23.3) Tg CO2eq yr^-1 and 9.7 (4.3-15.7) Tg CO2eq yr^-1, respectively for the Eastern Mediterranean region. These findings account for a ~17% and ~53% share of total European domain emissions as derived from our top-down assessment. The comparison with bottom-up emission inventories from these countries showed a good agreement for most compounds, except for large discrepancies for HFC-152a, for which UNFCCC reported emissions from Greece were more than a factor of four larger than our top-down estimate. The second campaign was performed at the high-mountain site Pic du Midi (French Pyrenees) in 2013-2015, where emissions related to densely populated areas in Spain and in France could be assessed. An inverse estimate of total HFC-134a and HFC-125 emissions from the Iberian Peninsula was relatively insensitive to utilisation of the additional measurements from Pic du Midi in addition to those from the three existing sites at Jungfraujoch, Mace Head and Mt. Cimone. However, a more detailed spatial allocation could be achieved when using the Pic du Midi observations.
Many halocarbons in the atmosphere are not monitored on a regular basis, as they have never been produced in large quantities and for this reason emissions are assumed to be negligible. Therefore, an additional key contribution of this thesis was the detection of previously unreported atmospheric halocarbons. It was possible to detect HCFC-31 (CH₂ClF) in the atmosphere for the first time world-wide. Northern hemispheric records showed that HCFC-31 concentrations were peaking in 2012 at ~170 ppq (parts-per-quadrillion, 10^-15) and decreasing afterwards. The measured mole fractions were used in a global 12-box model to estimate global peak emissions of 840 t yr^-1 (±90 t yr^-1) in 2011. The origin of the atmospheric HCFC-31 could be traced to emissions from the industrial production of HFC-32 (CH₂F₂). In this large-scale process hydrofluoric acid (HF) is used to exchange the chlorine atoms in dichloromethane (CH₂Cl₂) with fluorine. However, if the conversion is not complete, HCFC-31 is formed as a by-product. In recent years, global HFC-32 production has been rising continuously. Therefore, the recent decline in emissions of HCFC-31 could point to improvements in the production process.
This study shows the potential of the combination of measurements and meteorological transport models to act as an early warning tool for newly emitted gases and to estimate regional scale emissions in regions with limited information on emissions factors and activity data. In future, with application on the global scale, this could contribute to building trust between different countries related to their relative greenhouse gas emission inventories.