Abstract. The stratospheric contribution to tropospheric ozone (O3)
has been a subject of much debate in recent decades but is known to have an
important influence. Recent improvements in diagnostic and modelling tools
provide new evidence that the stratosphere has a much larger influence than
previously thought. This study aims to characterise the seasonal and
geographical distribution of tropospheric ozone, its variability, and its
changes and provide quantification of the stratospheric influence on these
measures. To this end, we evaluate hindcast specified-dynamics
chemistry–climate model (CCM) simulations from the European Centre for
Medium-Range Weather Forecasts – Hamburg (ECHAM)/Modular Earth Submodel
System (MESSy) Atmospheric Chemistry (EMAC) model and the Canadian Middle
Atmosphere Model (CMAM), as contributed to the International Global Atmospheric Chemistry – Stratosphere-troposphere Processes And their Role in
Climate (IGAC-SPARC) (IGAC–SPARC) Chemistry Climate Model
Initiative (CCMI) activity, together with satellite observations from the
Ozone Monitoring Instrument (OMI) and ozone-sonde profile measurements from
the World Ozone and Ultraviolet Radiation Data Centre (WOUDC) over a period
of concurrent data availability (2005–2010). An overall positive, seasonally
dependent bias in 1000–450 hPa (∼0–5.5 km) sub-column ozone is
found for EMAC, ranging from 2 to 8 Dobson units (DU), whereas CMAM is found
to be in closer agreement with the observations, although with substantial
seasonal and regional variation in the sign and magnitude of the bias (∼±4 DU). Although the application of OMI averaging kernels (AKs)
improves agreement with model estimates from both EMAC and CMAM as expected,
comparisons with ozone-sondes indicate a positive ozone bias in the lower
stratosphere in CMAM, together with a negative bias in the troposphere
resulting from a likely underestimation of photochemical ozone production.
This has ramifications for diagnosing the level of model–measurement
agreement. Model variability is found to be more similar in magnitude to that
implied from ozone-sondes in comparison with OMI, which has significantly
larger variability. Noting the overall consistency of the CCMs, the influence
of the model chemistry schemes and internal dynamics is discussed in relation
to the inter-model differences found. In particular, it is inferred that CMAM
simulates a faster and shallower Brewer–Dobson circulation (BDC) compared to
both EMAC and observational estimates, which has implications for the
distribution and magnitude of the downward flux of stratospheric ozone over
the most recent climatological period (1980–2010). Nonetheless, it is shown
that the stratospheric influence on tropospheric ozone is significant and is
estimated to exceed 50 % in the wintertime extratropics, even in the lower
troposphere. Finally, long-term changes in the CCM ozone tracers are
calculated for different seasons. An overall statistically significant
increase in tropospheric ozone is found across much of the world but
particularly in the Northern Hemisphere and in the middle to upper
troposphere, where the increase is on the order of 4–6 ppbv (5 %–10 %)
between 1980–1989 and 2001–2010. Our model study implies that attribution
from stratosphere–troposphere exchange (STE) to such ozone changes ranges
from 25 % to 30 % at the surface to as much as 50 %–80 % in the
upper troposphere–lower stratosphere (UTLS) across some regions of the
world, including western Eurasia, eastern North America, the South Pacific
and the southern Indian Ocean. These findings highlight the importance of a
well-resolved stratosphere in simulations of tropospheric ozone and its
implications for the radiative forcing, air quality and oxidation capacity of
the troposphere.
Trajectory model simulations of ozone and carbon monoxide in the Upper Troposphere and Lower Stratosphere (UTLS)