Satellite and ground-based ozone observing campaigns have both been operating for the last 40 years and there is now a vast amount of data, with which to diagnose variability and trends in stratospheric ozone. Despite this wealth of information, we cannot currently determine the status of long-term trends in ozone with confidence, and recent findings indicate that stratospheric ozone might not be recovering as expected from ozone losses last century due to decreases in the lower stratosphere (Ball et al., 2018). To build confidence, progress must be made in data merging to build robust long-term records.

This project aims to build a rigorous state-of-the-art statistical framework to overcome limitations currently associated with merging multiple instrument data sets with different observing characteristics and, ultimately, pave the way to a unified ozone data set for stratospheric and tropospheric ozone, with realistic uncertainty estimates, which can then be used to make robust conclusions about the state of atmospheric ozone.

An important reason as to why much uncertainty remains in ozone trend analysis is because standard techniques for merging different data sets, combined with a lack of reasonable uncertainty estimates for those data, introduce artefacts (e.g., drifts, discontinuities, and sampling errors); additionally, each observing platform has its own inherent systematics that need to be accounted for. These artefacts ultimately inhibit having complete confidence in the final merged data sets, which explains why there are now approximately 10 ozone composites comprised of different combinations of the available ozone instrument data, but whose analyses produce varying trends and uncertainties. Merging of ground-based and satellite data has proved particularly elusive, and no convincing approach has yet been put forward. The community has repeatedly stated that better methods are essential to making any further progress (e.g., Harris et al., 2015).

To address these problems, we formed an ISSI team that brings together experts in statistical methods and atmospheric science, from both space and ground-based instrumentation, to build the most advanced and comprehensive methodological framework for merging all ozone data. This combination of expertise is necessary to ensure this goal can be achieved: it requires a rigorous statistics perspective that incorporates known problems from instrumentation, and an understanding of atmospheric variability to avoid spurious conclusions about any apparent anomalies in the data sets themselves. Ultimately, it will lead to a general framework through which any essential climate variable (ECV) can be integrated from independent and complementary observing platforms into a unified long-term data record.

Ball, W. T., Alsing, J., Mortlock, D. J., et al.: Evidence for a continuous decline in lower stratospheric ozone offsetting ozone layer recovery, Atmospheric Chemistry & Physics, 18, 1379–1394, 10.5194/acp-18-1379-2018, 2018.

Harris, N. R. P., Hassler, B., Tummon, F., et al.: Past changes in the vertical distribution of ozone – Part 3: Analysis and interpretation of trends, Atmos. Chem. Phys., 15, 9965–9982, 10.5194/acp15-9965-2015, 2015.