|
ATMOSPHERIC
|
|
Meteorology of the 1996/97 Arctic Winter
The large interannual variability in the meteorology of the Arctic lower stratosphere ensures that each winter is unique and interesting in its own way. So far, the winter of 1996/97 is no exception. Following two winters (1994/95 and 1995/96) in which temperatures in the lower stratosphere were very cold for extended periods, 1996/97 has started by being very warm, with a very weak polar vortex.
Fig. 17 shows the minimum temperatures near 50 hPa in the polar lower stratosphere for the past 7 years. At this altitude rapid chlorine activation on polar stratospheric clouds (PSCs) starts around 195 K. The minimum temperatures in December 1996 were well above this PSC threshold and much warmer than the previous 6 years. (The use of UKMO analyses for 1996/97 versus ECMWF for the earlier years does not affect this comparison). This is in strong contrast to December 1995 when temperatures were near the ice point (around 188K). Subsequently, temperatures in these large- scale analyses dipped below 195K at this altitude during January. Fig. 18 shows a temperature map from the ECMWF analyses at 475K for January 9, 1997. There is only a small region, masked by the coarse contour interval, below 195K.
Associated with these very warm temperatures has been a very weak polar vortex. A corresponding PV map for January 9, 1997 is shown in Fig. 19. The PV gradients are still very weak and the vortex is not well defined. In Cambridge we are running a SLIMCAT 3D chemical transport model simulation for this winter in near real time using the UKMO analyses. Fig. 20 shows N2O from this SLIMCAT run also at 475K on January 9. This long-lived tracer illustrates the disturbed nature of the vortex and shows a tongue of vortex air being pulled out over Europe, with an apparent intrusion of mid-latitude air into the vortex.
The extent and duration of cold PSC temperatures will affect the degree of chlorine activation and ozone destruction. Although this winter has so far been very warm around 50 hPa, colder temperatures (relative to the PSC threshold) have occurred higher up, e.g. at 30 hPa. Also, it is worth noting that there is still time for the vortex to stabilise, temperatures to fall and significant ozone loss to occur in the lower stratosphere. Indeed, because of the increasing insolation, PSC occurrence in March, for example, would be much more important for O3 loss. This situation happened recently in 1993/94 (see Fig. 17) when, after a warm, disturbed January and February, the vortex strengthened, temperatures fell and significant processing occurred.
Martyn Chipperfield
(Cambridge, Chemistry)
Modelling Ozone Laminae in the lower stratosphere
The study of ozone laminae presented at the 1996 UGAMP annual conference (see Newsletter 15) has now been accepted for publication. In this study, observations of ozone laminae made in Feb. 1995 during the SESAME campaign were modelling using the contour advection technique. In particular, one case study around the 15th Feb. was chosen. About local noon at a number of European ozone sonde sites, laminae in the ozone profiles had been recorded. However, other European stations did not record the filament or saw laminae on different days. Of particular note is the observation of a ozone lamina over Greece on the 17th.
A number of contour advection runs were carried out to model the filament responsible for these observations. The runs were started from the 8th Feb. on the 490K isentropic surface. The T42 PV field was used to initialise the contours. Fig. 21 shows the situation on 15th Feb. A filament is clearly seen separated from the vortex and advecting across the European observing stations. The timing of the passage of the filament with the ozone laminae observations is remarkably good, agreeing to within 6 hrs. or so. The position of the filament also seems to agree well with the stations that observed and did not observe any laminae.
Fig. 22 shows the same filament but two days later. It has now moved further south and over the Mediterranean in good agreement with a sounding made over Athens at the same time.
The filament is a robust feature. A contour advection run started on the 10th shows a very similar feature.
Reference
Reid S.J., Rex M., von der Gathen P., Floisand I., Stordal F., Carver G.D., Reimer E., Kruger- Carstensen R., OÕConnor F.M., Braathen G.O., Murphy G., J. Wenger, C. Zerefos and C. Varotsos, A study of ozone laminae using quasi- isentropic trajectories, contour advection and photochemical trajectory model simulations, 1997, J. Atm. Chem. (SESAME special issue), accepted for publication.
Glenn Carver
(Cambridge, Chemistry)
Steve Reid
(NOAA, Boulder)
A Global Tropospheric Chemistry Model - TOMCAT
The global 3-D transport model, TOMCAT, is now running with a full treatment of tropospheric chemistry. The model has been run with methane oxidation chemistry in the first instance, using the ASAD integration package developed by Carver et al. (1997). Chemical schemes, including nonmethane hydrocarbons, have also been tested in the model. Photolysis rates are calculated off- line in the UGAMP 2-D model using a two-stream method which takes into account scattering by aerosols and clouds. Latitudinally and seasonally varying emissions of source gases (CH4, CO and NOx) are included in the model. Lightning emissions of NOx depend on the convective cloud top height derived from the Tiedtke convection scheme (Stockwell and Pyle, in preparation). Physical sinks for trace gases in the troposphere include loss by rainout (wet deposition) and dry deposition at the surface. In the model, wet deposition depends on the rate of change in the vertical water vapour gradient, trace gas solubility (Giorgi and Chameides, 1985) and the frequency of rainfall events (Jonson and Isaksen, 1991). The dry deposition flux of a trace gas varies according to surface type (grass, forest, water, snow/ice) and the vertical flux of momentum near the ground (Walcek et al., 1986).
The model is forced with winds and temperatures from ECMWF analyses. The humidity fields are used in the chemistry scheme in reactions involving water vapour. So far, the model has been run for two periods: July to September 1994 and January to March 1995. The model was initialised using chemical fields taken from the 2D model and a correction to give more realistic tropopause heights using a relationship between PV and ozone. The model was run at a horizontal resolution of T15 (approx. 7.5 degrees or ~800km) and a vertical resolution of 31 levels (approx. 0.6-1.0 km) and chemical fields stored every 6 hours.
The reason for simulating these two periods is that ozone data from the MOZAIC (Measurement of Ozone by Airbus In-service Aircraft) project is available which can be used to validate global chemistry models. MOZAIC makes continuous measurements of ozone and relative humidity worldwide (but mainly over the North Atlantic) on five Airbus A-340 passenger aircraft. The project is funded by the EU and Airbus Industrie (n.b. this data is unpublished; all enquiries should be directed to the coordinator of MOZAIC, Dr. Alain Marenco, Laboratoire d'Aerologie, 14 Avenue Edouard Belin, 31400, Toulouse, France.)
Two examples of this comparison are shown in Figs. 23 and 24. In Fig. 23 all available MOZAIC data and 6-hourly ozone fields from TOMCAT have been averaged between 10 and 11km for the period August to September 1994. In the MOZAIC data, the boundary between the troposphere and the stratosphere can clearly be seen over the North Atlantic where ozone concentrations exceed 100 ppbv. TOMCAT shows reasonable agreement with these results although the model ozone tropopause is somewhat lower than in the data. This may be due to the model being too diffusive resulting in smeared out gradients around the tropopause. This is shown more clearly in Fig. 24 which is a comparison between 2-day average model results and MOZAIC data over Frankfurt between the ground and 12km.
In regions over the tropical oceans there is good agreement between the model results and the data. Over South America, ozone concentrations are overestimated, possibly due to excessive biomass burning emissions of CO leading to additional photochemical production of ozone. The results from this study are being written up for a special section on MOZAIC in the J. Geophys. Res.
Kathy Law &
Paul-Henri Plantevin
(Cambridge, Chemistry)
Trajectory modelling of the EASE '96 campaign.
During the summer of 1996 the East Atlantic Summer Experiment 1996 (EASE '96) took place at Mace Head, on the remote west coast of Eire. EASE '96 is part of the NERC programme ACSOE (Atmospheric Chemistry Studies in the Oceanic Environment). The emphasis of EASE '96 was on improving our understanding of the oxidizing capacity of the North Atlantic. In order to achieve this aim, a comprehensive suite of measurements were made, both of physical parameters such as temperature, pressure, humidity etc., and chemical constituents of the air arriving at Mace Head. A large number of groups were involved in the campaign; UEA (Hydrocarbons, NOx, peroxides, NO3 and XO), Leeds (Hydrocarbons, OH and HO2), Birmingham (NOx, HNO3, HONO, aerosol), Bristol (O3, CO, CFCs, HCFCs, HFCs), Imperial (HCHO), ITE (PAN, NOx), Leicester (RO2) and Galway (Aerosol). Our group at Cambridge are involved in modelling studies based on the measurements made at Mace Head.
Modelling has concentrated on using the Lagrangian trajectory approach to investigate both the transport of long lived species and the chemistry of shorter lived species. Five day back trajectories ending at Mace Head were calculated based on ECMWF data and stored on the BADC. Four trajectories a day were supplied, thus a total of 144 trajectories described the campaign.
These trajectories can be used to drive the Cambridge Tropospheric Trajectory Model. The model is initialized from 2-D fields of long lived species (O3, NOx, CH4, CO etc.), and has a fairly comprehensive tropospheric chemistry scheme. Emissions of both biogenic and anthropogenic source gases are also included. By running the model along each trajectory a time series is calculated which may be compared with the actual measurements made at Mace Head.
The model agrees reasonably well with the measurements. During the campaign, Mace Head experienced different meteorological conditions which are reflected in the measurements. During anti-cyclonic weather high levels of ozone were measured at Mace Head due to photochemical production from anthropogenic NOx and hydrocarbons emitted over Europe. This event is captured by the model although the degree of ozone production is over estimated. Work to date has indicated the importance of isoprene (biogenically emitted gas) in controlling the degree of ozone production. When air from tropical regions is seen at Mace Head, low ozone concentrations are observed. This is not well reproduced by the model probably indicating problems with the initialization. It is hoped that initializing with 3D fields should help alleviate this problem.
The model reproduces many of the features seen during the campaign. Further studies during specific periods should allow for a greater understanding of the oxidation processes of importance in the North Atlantic.
Mathew Evans,
Kathy Law,
Dudley Shallcross,
Oliver Wild, John Pyle
(Cambridge, Chemistry)