370214 Uncertainty in Ozone Trend Detection

Monday, 13 January 2020
Hall B1 (Boston Convention and Exhibition Center)
Marianna Linz, Harvard University, Cambridge, MA; and J. L. Neu and P. Lin

With the exception of the recent uptick in CFC11 emissions, the production of ozone depleting substances (ODSs) has largely ceased, as mandated by the Montreal Protocol. The long lifetime of ODSs, however, means that equivalent effective stratospheric chlorine (EESC) has undergone a gradual decline since 1997 and will continue to decrease slowly throughout this century. While there are hints of recovery of the Antarctic ozone hole, the only region of the stratosphere in which statistically significant positive ozone trends have been measured is the tropical upper stratosphere, with only about half of the trend attributed to the decline in ODSs. Throughout the rest of the stratosphere, the expected recovery signal is small relative to natural variability. Lower stratospheric ozone, in fact, seems to have declined since 1997, which may partially explain the lack of significant trends in total column ozone. However, the observed lower stratospheric trends are highly uncertain, particularly in midlatitudes, and modeling studies highlight the very large natural variability in this region. Our ability to detect the trend forced by the slow decline of EESC is limited both by the length of the record and by the strong internal variability in the system. In order to determine the minimum trend we could detect with the existing observational record, we repeatedly sample the preindustrial control run of the GFDL ESM4 model with an approximation of our current observing system to get probability distributions of trends due to internal variability only. We examine each latitude and consider the upper stratosphere, lower stratosphere, and total column ozone all separately. Because of differences in internal variability in the different regions of the stratosphere, we see differences in the detectable trend—there is less internal variability in upper stratospheric ozone than lower stratospheric ozone, for example. We also sample the model with different lengths of observations to show how continuing and future observations will enable the detection of the trends due to the decrease in EESC.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner