 |
Stratospheric Processes And their Role in Climate
|
||||||||
| Home | Initiatives | Organisation | Publications | Meetings | Acronyms and Abbreviations | Useful Links |
|
Stratospheric Processes And their Role in Climate
|
||||||||
| Home | Initiatives | Organisation | Publications | Meetings | Acronyms and Abbreviations | Useful Links |
The Fifth European Workshop on Stratospheric Ozone was held in late September 1999 in Saint-Jean-de-Luz, France and was the successor of the ones held in Schliersee, Germany. The meeting was convened by CNRS, INSU, CNES and the European Commission (EC) Research DG-XII. It was attended by over 250 participants from across Europe, and in all 221 posters were presented. The European stratospheric research programme is supported by the EC Environment and Climate programme and by national agencies. Research results from the whole European stratospheric programme were presented and discussed at St-Jean-de-Luz, except for UV-B, aviation impact and climate/chemistry interaction for which overviews were presented based on recent results, assessments and meetings.
Each session comprised overview talks, talks on individual studies and posters. An important part of the meeting was the presentation and discussion of the results of the Third European Stratospheric Experiment on Ozone (THESEO), whose main field activities took place during the winter 1998/1999 and the first part of 1999 (see articles in SPARC newsletter 10 and newsletter 13). We discuss here some of the topics in which significant progress was reported. A complete account of the meeting will be presented in the Workshop proceedings which will be published in EC Air Pollution Research Report series in 2000 and which will contain extended abstracts from the work presented at St Jean de Luz as well as a summary of results from THESEO. The abstracts of the posters and talks presented during the Workshop are available on the European Ozone Research Coordinating Unit Web site (http://www.ozone-sec.ch.cam.ac.uk/). The authors wish to acknowledge the efforts and assistance of all involved scientists in the preparation of this article, particularly the THESEO core group and the individual project coordinators.
The two Arctic winters covered by THESEO offered the opportunity to extend the series of studies of ozone loss in the Arctic lower stratosphere (LS), which started in the late 1980s and continued throughout the 1990s. Since the first large-scale European campaign focussing on the ozone evolution in the Arctic regions (EASOE in 1991-1992, followed by SESAME in 1994-1995), eight Arctic winters have now been well investigated.
The two THESEO winters, 1997-1998 and 1998-1999, were two of the warmest winters in the 1990s and followed a series of winters which were unusually cold. The 1997-98 winter was the colder of the two with several periods of low temperatures between December and February when polar stratospheric clouds (PSCs) formed. There were only two cold periods in the 1998-99 winter, one at the start of December when PSCs were observed over northern Scandinavia and one in February where no observations of PSCs were made by the available instruments. In both winters the stratosphere was disturbed with no strong isolated vortex existing after December.
In conjunction with the more disturbed stratosphere and higher temperatures, ozone values over the Arctic were higher than in the previous few years. For example high latitude total ozone in March were comparable to the amounts seen in the 1980s, in contrast to the record low values seen in the previous two years. A similar, though less marked picture was observed over northern mid-latitudes.
Analyses of various stratospheric parameters (minimum T, possible area of PSC formation, frequency of stratospheric warmings, etc.) from the long term meteorological records which exist since the 1950s provide a good perspective on the better studied winters of the last decade. This long-term view is also important in understanding the evolution of each individual winter. For example, the first major mid-winter warming since 1990/91 occurred in December 1998, the longest gap between such events in the existing 41-year record (a four-year gap was the previous record). The significance of this longer gap remains unclear; is it a natural fluctuation or part of long-term trend?
Now that the main qualitative features of chemical depletion of ozone in the Arctic regions are understood and reproduced by the models, studies concentrate on interannual variability and respective roles of the dynamics and of the chemistry in ozone depletion in the Northern hemisphere. These studies will help to provide a better quantitative understanding of ozone loss. Stratospheric models currently tend to calculate too little loss in comparison to the ozone loss derived from measurements, though in some winters (notably 1998/99) the models calculate larger losses than determined empirically. Only a correct understanding of the interplay between dynamics and chemistry will allow more accurate calculations of the chemical ozone loss to be made in the models.
During the two THESEO winters, 1997-1998 and 1998-1999, there were only short episodes of temperatures about or lower than the PSC formation temperatures in the Northern hemisphere. However there was some ozone loss calculated at the end of each winter in the Northern hemisphere by the different methods, based on ground-based, balloon, satellite measurements and 3D CTM outputs, or using a combination of those techniques. These several independent methods are now well validated and show a good qualitative agreement. Another, purely modelling approach presented is based on tracer experiments made with SLIMCAT. A series of tracers is initialised in several latitude bands. These model experiments can separate the calculated in-situ ozone loss from the influence of air transport from other latitude regions. They confirm that a lot of mixing and transport of activated and/or ozone-depleted air occurred during the two THESEO winters.
The chemical ozone loss occurred under threshold conditions (temperatures around those at which fast chemical loss can take place) which have not been extensively studied before in the 1990s. One notable example which will be a good test of our understanding of cold aerosol chemistry was the period in early February 1999 temperatures hovered around the PSC existence temperature. No PSCs were observed by the available instruments and a number of measurements of active halogen species such as BrO, ClO, OClO, NO2 and NO3 were made in this period. These should provide good constraints to calculations and this event offers a good opportunity to investigate the reactions on cold aerosol and PSCs in marginal temperature conditions. This is important as the 3D models calculated that appreciable ozone loss occurred after this cold period.
The causes of middle latitude ozone decline remain controversial and their study was the major aim within THESEO. It has been argued by some that much of the loss occurs in-situ while others have suggested that polar processes have a significant influence on the decline of ozone in mid-latitudes. A number of studies within THESEO, both observational and modelling, are making an important contribution to the debate.
A major effort has been made to develop empirical techniques capable of measuring chemically induced changes in ozone over mid-latitudes. These changes are hard to detect as they usually occur at much slower rates than in the Arctic, and distinguishing the effects of chemistry and dynamics is much trickier. However promising starts have been made with a couple of approaches whose preliminary results show that chemical ozone changes did take place over mid-latitudes, apparently unrelated to vortex processes, and that the ozone loss occurs faster at lower temperatures in qualitative agreement with current chemical understanding. When confirmed these studies will represent the first direct observational evidence of in-situ chemical ozone changes outside the vortex.
Chemical transport models have been used to examine the connection between polar and middle latitude loss. The winter of 1998/99, the second THESEO winter, has proved to be especially interesting. Temperatures in the LS were higher than average. There was, however, some activation of chlorine inside the polar vortex although outside the vortex synoptic scale temperatures were too high for activation. There were major dynamical disturbances to the vortex in mid-December and in late February. Model calculations show that there was a considerable exchange of air between the polar vortex and middle latitudes associated with these events. There was only a relatively small amount of chlorine-induced loss but this was manifested in the model throughout middle and high latitudes. Model calculated ozone losses did not show a large latitudinal gradient. Clearly, in this winter at least there was a considerable connection between ozone loss in middle and high latitudes.
Interesting comparisons can be made with the modelled ozone losses in earlier winters, including the first THESEO winter, 1997/98. For example, the vortex of 1996/97 was stable and long-lasting with low temperatures again confined inside the vortex. Model calculations indicate that while there was substantial ozone loss inside the vortex there was only a modest decline of ozone outside the vortex. This is in contrast to the model results for more dynamically disturbed winters which show a larger loss outside the vortex, and less contrast with the in-vortex depletion. Thus modelling in THESEO, including studies of earlier winters in the 1990s, has shown that there is a significant and important connection between polar and middle latitude ozone loss. A cold, stable vortex favours large polar loss with little connection with middle latitudes but in more disturbed winters polar loss, and polar processing, contribute to middle latitude decline.
The results from THESEO thus confirm that both in-situ and polar processes contribute to the halogen-induced ozone loss over mid-latitudes. The relative contributions vary from year to year and depend on factors such as the stratospheric temperatures over mid-latitudes and the strength and temperatures of the Arctic vortex.
A number of chemical studies were presented at the meeting, based on laboratory and modelling studies as well as field observations. One issue, which has received particular attention in the last couple of years, is the role of bromine in the chemical destruction of ozone.
Bromine monoxide was extensively monitored during THESEO by means of UV-visible spectrometers operated from the ground, balloon and satellite (ERS-2/GOME). Altogether these observations form a data set of unprecedented completeness, which have been used to test our current understanding of the stratospheric bromine chemistry at various latitudes, seasons, and time of the day. Comparisons with calculations by 3D chemical transport models reveal a good overall degree of agreement between modelled and observed stratospheric BrO amounts, and their variations. For example, the latitudinal and seasonal variability of BrO columns and profiles are well reproduced. Balloon profiles measurements have shown that bromine is activated (about 60% in BrO during daytime) at all latitudes and seasons. Most of column changes previously reported from ground or satellite observations are therefore due to transport (vertical displacement of the stratosphere) and not to chemistry. The variability is well reproduced by 3D CTM models, as well as signatures of chlorine activation occasionally observed in twilight BrO columns and explained in the model by the formation of the photolabile BrCl. Finally there is now agreement with respect to the total amount of bromine derived from source gas measurements and the value inferred from BrO measurements, with the inclusion of short-lived compounds such as bromoform give a total of about 20pptv.
The good overall agreement gives us confidence that current models can be used for an assessment of the role of bromine in mid-latitude ozone loss. However some discrepancies between model and observations still exist which require further study, e.g. the behaviour of morning to evening column amounts ratios not well reproduced at all ground-based stations, and the partitioning between bromine species is controlled by NO2 which is largely underestimated by the models in the LS.
The observed NOx concentrations in the LS are still significantly larger than those modelled. Recent revisions of laboratory data of chemical reaction rates have reduced the differences in modelled and measured concentrations as well as in the delay in the NOx recovery at spring, but there are still significant disagreements. Models still fail to capture the presence of NOx at spring in the lowermost stratosphere as in the model there is fast conversion of HNO3 on cold aerosol in contrast to observations. Since the discrepancy is similar every year in the long series of profiles started in 1992 when the aerosol loading was 10 times larger, the likely explanation would be in the NOx gas phase chemistry at lower temperature which requires to be re-visited.
Indirect evidence of chlorine activation has been obtained from aircraft measurements in the lowermost stratosphere by the observation of anomalously enhanced CO/C2H6 ratios. The measurements also confirmed the presence of subvisible cirrus above the tropopause, suggesting that local chlorine activation has taken place on cirrus crystals. Furthermore, a chemistry box-model and a three-dimensional chemistry-transport model were used to investigate the photochemical effects of enhanced ice particle surface areas in the lowermost stratosphere. It appears that ice particles in the lowermost stratosphere can cause chlorine activation and chemical O3 loss during most of the year, not only limited to the cold seasons (winter and early spring). Low temperatures thus do not seem to be a prerequisite for chlorine activation, as long as a significant reactive particle surface is present. The combination of heterogeneous chlorine activation with UV light intensity may play a significant role in local chemical O3 loss in the (sub)tropical lowermost stratosphere.
The relatively high temperatures in the Arctic regions during the 1997/98 and 1998/99 winters were associated with a very disturbed vortex, unstable and often displaced from the high latitudes. These disturbed conditions resulted in extensive exchange of air between the vortex and mid-latitudes throughout the two winters. Specific studies investigated mixing processes between the different latitude regions (and in the UT/LS region) and their impact on mid-latitude ozone evolution. Results and analyses from studies of polar erosion, mixing of air from the sub-tropical UT/LS were presented. They were based on a combination of aircraft measurements in the LS and ground-based and ozonesonde measurements made on alert inside the filamented airmasses. New numerical tools were developed in parallel, efficiently supporting the analysis of the data. A good set of case studies during the two THESEO winters has been assembled as well as the results from the STREAM 98 campaign in Canada in summer 1998.
In-situ tracer measurements with balloon-borne samplers were made in winter and spring 1998/99 in order to investigate the extent of mixing between polar vortex air and mid-latitude air. The tracer data are being compared with similar measurements which were carried out in winter and spring 1996/97, when the polar vortex was extraordinarily robust in March and April persisting through early May. Inside the winter Arctic polar vortex "old" air masses (in terms of mean age and photochemical exposure time) descended from the upper stratosphere/lower mesosphere along with diabatic cooling. After the break-up of the polar vortex, blobs of vortex air can maintain their integrity for up to one or two months and can be advected to lower latitudes, as observed in measurements made at the end of June in 1997 over Gap (Southern France, 44¡N).
The results of studies on large-scale transport and stratospheric ozone as calculated by CTMs were presented. The use of tracer relations allowed to investigate the age of air and how well the mixing is simulated by the models. Comparisons with measurements (long-lived tracers) provide a good test of model transport (e.g. descent). New theoretical approaches are being developed (such as effective diffusivity to investigate mixing processes).
Several studies have concentrated on the processes specific to the tropical regions. The results of a large-field campaign (APE-THESEO) involving the high-flying M55 Geophysica and the DLR Falcon revealed the existence of a very cold tropopause and of a very broad ozone tropopause over the west Indian Ocean. A report of activities within APE-THESEO is presented elsewhere in this Newsletter.
Satellite cloud data and real-time aerosol lidar data from the DLR Falcon aircraft were used to guide the M55 Geophysica aircraft, which thus was able to penetrate the thunder cloud anvils and sub-visible cirrus so the in-situ trace gas and particle measurements were made to complement the remote measurements from the lidar. Measurements from both aircraft showed the existence of very thin sub-visible cirrus which extended over considerable distances. This was something of a puzzle as the observed clouds existed across too large a temperature range for the cirrus to consist of ice particles. Preliminary analyses of measurements from a new instrument indicate the presence of HNO3 in the condensed phase. Model calculations based on measurements from several instruments and on laboratory data indicate that the particles are made of nitric acid trihydrate, the normally polar particles.
For the first time a large amount of in-situ tracer observations was obtained in the tropical UT/LS between the equator and 20¡S during the deployment of the M55 Geophysica aircraft in February/March 1999 from the Seychelles (5¡S). A newly developed in-situ gas chromatograph provided measurements of a variety of long-lived tracers (N2O, F12, F11, H-1211, SF6) along with high resolution CO2 measurements by an integrated infrared analyser. The vertical profiles and tracer correlations with ozone in the tropical LS reveal that most of the air sampled has slowly ascended from the tropical tropopause largely, though not completely, in isolation from midlatitude air. In one flight, however, a recent intrusion of southern midlatitude air to 11¡S at 19km altitude is apparent. In the upper troposphere (UT) a weak north-south gradient is observed for tracers with significant interhemispheric differences at the surface, suggesting that air masses from the two hemispheres ascending within the Inter-Tropical Convergence Zone (ITCZ) have not completely mixed by the time they reach the stratosphere.
A field of growing importance is the interaction between climate change and the stratosphere with a number of chemical and dynamic factors coming into play. Several long-term changes in the stratosphere have been detected, associated (though possibly not uniquely) with the ozone changes. A long-term negative trend in the temperatures has been calculated at high latitudes. A consequently stronger vortex would then extend longer in spring. Clearly the variability of the Arctic stratosphere (as evidenced in the last few years) makes it hard to distinguish trends from natural variability. However a colder vortex would lead to greater ozone loss for a given halogen loading and so delay the anticipated ozone recovery as halogen levels fall. This effect may be enhanced by microphysical processes such as greater sedimentation with consequently longer periods of ozone loss.
Changes in N2O columns measured over the Jungfraujoch since the 1950s coupled with in-situ N2O measurements indicate an increase in tropopause height consistent with the increase measured since the late 1960s at Hohenpeissenberg. This increase in tropopause height has occurred over much of Europe and has affected ozone amounts over Europe. An analysis of total ozone measurements at Hohenpeissenberg indicates that part of the ozone trend observed there is related to changes in tropospheric circulation related to the Polar-Eurasia teleconnection pattern; and an analysis of ozone measurements from Switzerland also reveals a significant dynamic contribution to the observed ozone trend, although in this case the dynamic connection is ascribed to the North Atlantic Oscillation. Further work is needed to understand whether this is a regional phenomenon or whether there are implications at the hemispheric and global scales and whether it results from natural variability or climate change.
The results from THESEO presented here are the product of on-going research and further analysis is being made, particularly of the results at mid-latitudes where the interplay of chemistry and dynamics is subtle, yet important. A number of the projects in THESEO have been extended to include the 1999/2000 winter. Some of the research in THESEO required cold conditions (e.g. the measurements of the chemical composition of PSCs) and the extension means that these can be addressed further. A new project is also being supported through the EC FP5 programme. These projects comprise THESEO 2000. A close collaboration has been forged between THESEO 2000 and the NASA SOLVE campaign which is studying Arctic ozone loss in this winter. The results from these two campaigns and the further results from THESEO will greatly add to our understanding of ozone loss at mid and high latitudes in the northern hemisphere.