European F-gas emissions evaluation

Objectives and ambitions

F-gases, here referring to hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), NF3 and SF6, are potent GHGs with global warming potentials (GWP) up to several thousands of times that of CO2. F-gas emissions are rapidly evolving due to changes in production and use brought about by the Kigali Amendment to the Montreal Protocol and regional regulations. The ambitious European F-gas regulation (current version adopted in 2015) pledges to cut EU F-gas emissions by two-thirds of 2014 levels by 2030. Currently, top-down quantification of European F-gas emissions is limited to only a few countries with a sufficient density of observations: the UK, Switzerland, and northern Italy. Our previous work in the UK showed significant over-reporting of the country’s HFC-134a, which resulted in a subsequent revision of the UK HFC inventory (Manning et al., 2021; Say et al., 2016). In PARIS, this successful and proven approach will be refined and extended to cover a significant part of the European continent by adding new F-gas observations and expanding top-down F-gas estimation to eight countries. In tandem, we will advance spatial disaggregation of bottom-up estimates using unique industry knowledge of the F-gas sector. This powerful combination of top-down and bottom-up information will allow us to better inform inventory compilers of the accuracy of F-gas inventory models.


Three major limitations in our understanding of F-gas fluxes in Europe are addressed in PARIS: 1) limited measurement network, 2) poor understanding of the spatial and temporal distribution of emissions and, 3) the wide range of emission factors used by different countries in UNFCCC reporting.

F-gas measurement sites – Existing high-frequency measurement stations are shown as black dots (AGAGE and affiliated). New high frequency measurements will be installed at Taunus (TOB). New flask sampling is proposed at the locations indicated by a red star, at Birkenes (BIR), Cabauw (CBW), Hegyhátsál (HUN) and a site in southern Italy (TBD). Countries shown in light grey are those where national emissions inferences are likely to be possible, and focus countries are shown in dark grey.

New continuous measurements of a wide range of F-gases (here hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), NF3 and SF6) starting in February 2023 will be provided from the Taunus Observatory in Germany. In addition,  PARIS will add new F-gas observations in Norway, Hungary, The Netherlands and in southern Italy to cover a significant part of the European continent and expand top-down F-gas estimation to 8 countries. In addition, PARIS will advance spatial disaggregation of bottom-up estimates using unique industry knowledge of the F-gas sector, thus creating a powerful combination of top-down and bottom-up information.

Our focus is on the interpretation of the most important F-gas climate forcers in Europe: HFC-134a, -143a, -125, -32, -227ea, -365mfc, -152a, HFC-23, HFC-245fa, HFC-43-10mee, PFC-14, -116, -218, -318, NF3, SF6. To determine an optimal flask sampling strategy, observing system simulation experiments (OSSEs) will be performed. Compared to the more abundant GHGs, spatially resolved bottom-up estimates of F-gas emissions are very sparse and poorly documented. Therefore, our aim is to develop a new spatially resolved estimate of European F-gas fluxes. Furthermore, we will derive top-down F-gas emissions for the above mentioned countries using a set of inverse modelling frameworks. Emissions will be derived for the period 2010–2023 based on the current observational capabilities and, for 2024 and 2025, based on the expanded observations from PARIS. The so gained new top-down and bottom-up emission estimates will be included in the work of reconciling national bottom-up / top-down estimates.


Establish sampling strategy – Observing System Simulation Experiments (OSSEs) using the modelled mole fractions will be performed to design optimal sampling frequency and site operators will be informed of target sampling frequency (approximately 4 times per week averaged over each year).

F-gas flask sampling – Sampling at all sites will be conducted in accordance with the best possible strategy identified in the task “Establish sampling strategy”. Sampling will be carried out for at least two years at each site (2024 and 2025).

Continuous F-gas measurements – Continuous measurements from Taunus (Germany) will be implemented by employing a GC-MS Medusa system following AGAGE protocols. Data from four existing continuous measurement stations will also be included in this project: Monte Cimone, Italy (UNIURB), Zeppelin, Norway (NILU), Mace Head, Ireland and Tacolneston, UK (both UNIVBRIS). In addition, data from other AGAGE stations will be available for the project.

Improved bottom-up a priori emissions – Bottom-up emission estimates will be compiled by Oekorecherche (ORB), a specialised small German company. Due to the in-depth knowledge of the location of European F-gas production facilities, ORB will provide estimates of “hot-spots” of F-gas emissions through Europe with a focus particularly on the focus countries of Hungary, U.K., Germany, Italy, Norway, Switzerland, Ireland and the Netherlands, where improved observational data will be available.

F-gas flux estimation – The new observations from the tasks “F-gas flask sampling” and “Continuous F-gas measurements” will be assimilated in addition to existing measurements and national flux estimates will be derived at seasonal to annual resolution. Special focus will be given to improved a-priori emission estimates gained in the task “Improved bottom-up a priori emissions” and their impact on a-posteriori emissions will be assessed. 

Comparison of top-down with UNFCCC reports and inventory models: Top-down and reported emissions of the F-gases will be compiled for countries constrained by the observational network. In the focus countries, these estimates will be compared annually in an annex to the National Inventory Report. Discrepancies will be discussed considering best available error estimates for both bottom-up and top-down approaches and using knowledge developed in task.