Body text
Following the discovery of rapid ozone depletion over Antarctica and long-term decreases at mid-latitudes, the 1987 Montreal Protocol was established to phase out ozone-depleting substances and to mandate regular scientific assessments of the ozone layer.
The LOTUS research community, coordinated by BIRA-IASB, contributes to these assessments and has provided conclusive evidence of ozone recovery in the upper stratosphere since 2000. However, trends in the lower stratosphere remain highly uncertain due to large measurement uncertainties and strong natural variability, particularly that linked to the El Niño–Southern Oscillation (ENSO).
Stratospheric Ozone and El Niño Southern Oscillation
ENSO is a naturally occurring cycle of warming and cooling of the tropical Pacific Ocean that recurs every two to seven years. It exerts a strong influence on global oceanic and atmospheric circulation, affecting both the troposphere and the stratosphere (Figure 2).
Observations show that ozone concentrations in the tropical lower stratosphere decrease during El Niño events due to faster transport of ozone-poor tropospheric air into the stratosphere. Consequently, accurate quantification of long-term ozone trends requires careful consideration of this interannual variability, especially in the lower stratosphere.
Estimating ozone response time to ENSO
Long-term ozone trends are typically derived by regressing observed ozone time series against known sources of interannual variability, including ENSO. Traditionally, these analyses assume that stratospheric ozone concentrations are modulated instantaneously by ENSO events, though this has not been quantitatively verified using observations.
To address this gap, as part of the BRAIN-be 2.0 project TAPIOWCA, researchers at BIRA-IASB conducted a comprehensive analysis using three multi-satellite data records extensively demonstrated as long-term references to quantify the response time of stratospheric ozone to ENSO.
Findings and implications
Our results show consistent evidence that tropical lower stratospheric ozone responds to ENSO with a lag of approximately 2–6 months (Figure 3).
Longer response times of 3–9 months occur in the tropical and mid-latitude upper stratosphere, as well as in the mid-latitude lower stratosphere, though with considerably lower certainty. Accounting for these response times in regression analyses changes ozone trend estimates by less than 0.5 % per decade and reduces their associated uncertainties by a comparable amount. These adjustments are small relative to typical trend magnitudes (1–2% per decade) and uncertainties (2–3% per decade) in the lower stratosphere.
In conclusion, accounting for the non-instantaneous nature of the ENSO response improves regression accuracy, and we now have better confidence that it has a fairly minor influence on the conclusions of long-term stratospheric ozone trend assessments.
