Stratospheric impact on tropospheric ozone variability and trends: 1990-2009
Résumé
We evaluate the influence of stratospheric ozone on the interannual variability and trends in tropospheric ozone from 30-90° N between 1990 and 2009 using ozone measurements and a global chemical transport model (the Community Atmospheric Model with chemistry) with input meteorology from the National Center for Environmental Prediction. The model simulation uses constant interannual emissions. Both the model and measurements indicate that on large spatial scales stratospheric interannual ozone variability drives significant tropospheric variability and contributes to long-term tropospheric ozone trends. To diagnose the measured variability we utilized measurements from ozonesondes and the Measurements of OZone and water vapour by in-service Airbus airCraft programme (MOZAIC) north of 30° N. We identify a regionally robust 150 hPa ozone signal from measurements over Canadian, Northern European and Central European regions and at 500 hPa over Canadian, Northern European and Eastern US regions. Averaged over these regions, the 150 hPa interannual ozone variability explains 69 % of the interannual variability at 500 hPa. The simulated stratospheric signal explains 81 % of the simulated variability over these same regions. Simulated and measured ozone are significantly correlated over these regions and the simulation suggests that the ozone record over these regions is representative of the overall hemispheric 500 hPa ozone record from 30-90° N. The measured 500 hPa trends averaged over these three regions between 1990 and 2000 and 1990 and 2009 are 0.73 (±0.51) ppbv yr−1 and 0.27 (±0.19) ppbv yr−1, respectively. The simulated trends in 1990-2000 and 1990-2009 are 0.29±0.10 ppbv yr−1 and 0.13±0.05 ppbv yr−1, respectively; however, these trends are substantially larger when the model is sampled for missing data exactly as the measurements are. Simulated stratospheric ozone accounts for 79 % of the simulated 500 hPa trend between 1990 and 2000 and 100 % of the simulated trend between 1990 and 2009. Due to the importance of local meteorology and emissions at the surface it is difficult to isolate the stratospheric component of measured surface ozone variability. Overall when averaged between 30-90° N simulated surface interannual ozone trends are 0.18 ppbv yr−1 and 0.07 ppbv yr−1 between 1990 and 1999, and between 1990 and 2009, respectively. We have identified a number of surface sites where the measured interannual ozone variability is correlated with the 150 hPa ozone signal. Most notably these sites include the high mountain sites over Europe and Macehead, Ireland. Over Macehead the measured 150 hPa ozone signal explains 40 % of the interannual variability of the unfiltered measured ozone record. The simulated and measured ozone are highly correlated over Macehead. The Macehead measured and simulated unfiltered ozone trends between 1990 and 2000 are 0.28 (±0.33) and 0.17 (±0.13) ppbv yr−1 respectively; between 1990 and 2009 the measured and simulated trends are 0.18 (±0.11) and 0.08 (±0.06) ppbv yr−1, respectively. Increases in the simulated stratospheric ozone component accounts for 53 % and 75 % of the overall modeled trend for the two periods at Macehead.
