Review of AGU Fall Meeting Session: New Insights into Stratospheric Chemistry, Dynamics, and Transport

Emily Shuckburgh, LMD/ENS, Paris, France (eshuck@ravel.ens.fr)
Warwick Norton, Oxford University, Oxford, UK (wan@atm.ox.ac.uk)

Introduction

This session, convened by Emily Shuckburgh and Warwick Norton, presented recent advances in middle atmosphere chemistry, dynamics and transport. The session included 21 talks and 51 posters. The talks were broadly divided into four areas: (i) the effects of climate change on the stratosphere, (ii) new results and diagnostics of tracer transport, (iii) analysis of recent field campaigns and satellite measurements, with particular focus on the results of SOLVE-THESEO 2000, and (iv) understanding the distribution and transport of water vapour.

Climate Change

Elisa Manzini (invited) presented results of simulations using the MA-ECHAM4-CHEM model to evaluate temperature and circulation changes related to changes in atmospheric composition from 1960 to 2000. With respect to the 1960 simulation, the 2000 simulation showed significant average cooling (~6 K) in March [30-10 hPa], strengthening of the zonal winds (~8 m/s), and local decrease (~10%) in downwelling, mainly due to a change in the contribution from planetary waves. In contrast, in the mesosphere, downwelling is enhanced (~20%); the contribution from gravity waves is twice that from planetary waves. Christoph Brühl showed that the MA-ECHAM4-CHEM model, which includes interactive photochemistry, could reproduce the main features of global temperature and ozone distributions in the sixties and the nineties.
David Rind used the GISS Global Climate/Middle Atmosphere Model to examine the effect on the stratospheric circulation and tracer transport of doubled CO₂ and two alternative sets of SSTs, one with greater low latitude and overall warming. While both results in an intensified subtropical circulation in the lower stratosphere, they have opposite effects on the high latitude residual circulation and the Arctic Oscillation phase.
Piers Forster (invited) used the latest SPARC water vapour trends (assuming the NH mid-latitude trend to be representative of the global trend) to model stratospheric temperature changes and to calculate radiative forcing. He finds significant sensitivity in the model response to the imposed background water vapour concentration, but, with the best-guess background and trend applied in model, he obtains a 0.8 K global cooling of lower stratosphere for the 1980-2000 trend and a 1.8 K cooling for 1960-2000 trend, equating to radiative forcings of 0.3 Wm² over last 20 years or 0.6 over last 40 years. These forcings are large (75% of the CO₂ forcing or 40% of the well-mixed greenhouse gas forcing) and hence potentially important.
Betsy Weatherhead discussed optimal methods of detecting the gradual recovery in the total column ozone record, expected to be only a few Dobson Units per decade over the next 50 years. She explained that variations in the seasonal and altitudinal distributions of ozone suggest that evidence of recovery may be easier to detect during the local springtime, or at a single vertical layer in the atmosphere.
John Farrara used the UCLA GCM to examine the response to reduced stratospheric ozone, which may arise from reduced solar activity. He suggested an amplification mechanism, which through coupling by stratospheric annular modes could contribute to the little ice age period.

Tracer Transport

Bob Hudson (invited) showed that the total ozone distribution can be separated into four "meteorological regimes" divided by the subtropical upper troposphere front; the polar front and the polar vortex, with the distribution often being rather narrow in the first three, mid-latitude, regimes. He showed that the mean total ozone value within each of these mid-latitude regimes is relatively constant from year to year, suggesting that it is the relative areas of the three regimes that determine the mean mid-latitude (20 to 60°N) total ozone. He proposed that most of the northern mid-latitude ozone trend between 1978 and 1992 is in fact a consequence of a northward movement of the mean latitude of both the subtropical and polar fronts.
Anne Douglass examined diagnostics of tropical transport in CTM simulations driven by winds from the Finite Volume CCM and the DAO analyses. Signatures of interannual differences related to the tropical winds and extratropical waves were presented to understand differences between the two model transports and the transport inferred from the observations.
Fred Moore used in situ measurements of long-lived trace gases to calculate a distribution of maximum path heights over which air has travelled in the stratosphere. The measured maximum path height distribution is a transport diagnostic, which is independent of measured mean age or age distributions and thus provides an additional check on model derived stratospheric transport.
Bernard Legras demonstrated how ideas and concepts from the theory of dynamical systems could be used to investigate the transport characteristics of the stratospheric polar vortex. He used "finite-size Lyapunov exponents" to identify so-called stable and unstable manifolds in the vicinity of the vortex, and showed how these structures are associated with transport of particles.

Field Campaigns and Satellite Measurements

Megan Northway showed results from SOLVE-THESEO, which indicate widespread denitrification near 40% between 20-30 km in the polar vortex. To explain this, particle populations must have molecular HNO₃ flux values between 10⁹ and 10¹⁰ molecules cm^-3 km day^-1, these flux values must be achieved near 20 km for a total of several days, particle sizes need to be 5-20 mm diameter, and number concentrations less than 0.1 cm^-3.
Paul Wennberg (invited) and co-authors illustrated how the well-known formation of NAT particles in localised lee-wave mountain ice clouds can produce a low number density of large particles in the lower stratosphere. They showed, using a simplified model, that this mechanism for nucleation of NAT particles is quite consistent with observations made during SOLVE-THESEO. Other more speculative nucleation mechanisms (e.g. selective homogeneous or heterogeneous nucleation) may not be required to explain denitrification of the polar vortex. [Manuscript now in press at GRL, Dhaniyala, McKinney, and Wennberg, 2002].
Steve Eckerman explained that global forecast models used during SOLVE-THESEO indicated that the very coldest intravortex regions were small “mesoscale cold pools” occurring recurrently over the East Coast of Greenland, embedded within a broader synoptic cold pool region. He showed, based on archived forecasts and analyses from various global models, that these features are not artefacts of sigma coordinate models induced by the Greenland coastal topography, but are produced by a long-wavelength mountain waves, launched by eastward flow over the Greenland ice shelf.
Paul Konopka presented the results of high-resolution ClaMS model simulations of the chemical composition of the Arctic stratosphere for late spring 2000, and discussed the impact of mixing on ozone depletion. He showed that the deactivation of ClO_x within the air masses originating from the polar vortex occurred, even after the vortex break-up, mainly through the photochemical NO_x production from HNO₃ without a significant contribution from mixing with NO_x-rich mid-latitude air. He also showed the mid-latitude ozone deficit in late spring was dominated by dilution by polar air and, until mid of April, by ClO_x-induced ozone destruction within the vortex remnants.
Hideaki Nakajima presented tracer data from the Improved Limb Atmospheric Spectrometer (ILAS) on board the ADEOS satellite. When plotted in equivalent latitude space, the data nicely showed the vortex edge and subtropical barriers.
Peter Preusse analysed the high spatial resolution stratospheric temperature data sampled by CRISTA for gravity waves. Deduced gravity wave amplitudes were correlated to a tropopause-height/ convection proxy inferred from CRISTA water vapour and cloud top heights. This was thought to be due to two reasons: deep convection directly excites gravity waves as illustrated by super typhoon Winnie, and the vanishing Coriolis force at the equator allows the existence of low frequency gravity waves with very long horizontal wavelengths enhancing gravity wave energy in the tropics.

Water Vapour

Steve Sherwood (invited) showed there exists a close coherence between relative humidity variations just below the tropical tropopause, and moisture variations later at higher levels, from which temperature effects have been removed. He demonstrated that there is a highly significant correlation between the relative humidity of air crossing the tropopause and the effective diameter of ice crystals in cumulonimbus cloud tops, consistent with a physical explanation based on crystal fall speeds and evaporation rates. There is a likely link of crystal sizes to droplet nucleation by biomass burning particles at earlier cloud stages, and this is a likely contributor to the upward trends in stratospheric water, however, absolute calibration of trends in crystal diameter appears impossible at this time.
Koji Yamazaki used an AGCM simulation forced by observed SSTs to examine the downward time-mean vertical motion near the tropopause over Indonesia. A heat budget calculation showed that cold advection from the east and reduced longwave heating balances the warming by downward motion. Sensitivity calculations of the cloud height and temperature profiles showed that the high cloud suppresses the longwave radiative warming around the tropopause region.
Daniel Kirk-Davidoff presented preliminary results from the NASA-funded CWVCS airborne field mission, based in Costa Rica, to study the mechanisms controlling the distribution and transport of water vapour in the tropical upper tropopause and lower stratosphere. Measurements were made, including ozone, water vapour, total water and methane, in and far from regions heavily influenced by deep convection. A wide variety of very low-temperature cirrus clouds were intercepted, and tropopause temperatures as low as 185 K were encountered.
Zhiming Kuang showed observations of the isotopic ratio of water vapour in the tropical upper troposphere and lower stratosphere from ATMOS and the MkIV balloon flights, and noted that the isotopic ratios were not inconsistent with dehydration due to condensation down to water vapour mixing ratios of roughly 10 ppmv. Further drying of air down to 3-5 ppmv, which is typical of air entering the tropical stratosphere, was postulated to be due to some form of mixing, since the isotopic ratios were inconsistent with condensation as the exclusive drying mechanism below 10 ppmv.
Ghassan Taha explained that recent improvements to the SAGE II retrieval algorithm have resulted in enhancements in the vertical structure of the observed profiles. He discussed the latitudinal distribution and the climatology/interannual variation of water vapour obtained, and compared these to measurements from HALOE and MLS.