Arctic Winter 1995/96

The Arctic Winter of 1995/96 is well underway. As with all winters, the meteorological variability of the Arctic lower stratosphere will no doubt ensure that this winter is interesting in its own way. Moreover, this winter has already seen extremely cold temperatures in the Arctic lower stratosphere. According to the Berlin analyses (from B. Naujokat et al.) minimum temperatures during January at 30 hPa and 50 hPa reached 180-187 and 185-190 respectively. These correspond to record low temperatures. Two extreme cold periods occurred from 10 - 16 January and from 22 - 24 January when radiosondes measured record low temperatures: 178 K at Ny Alesund on 10 January at 28 hPa and 180 K at Bodo on 22 January at 21 hPa. At the end of the month, temperatures slightly increased due to a minor warming in the upper stratosphere. Figure 10 shows the minimum temperatures at 50hPa north of 60N from the ECMWF analyses so far this winter (thick line) compared to the 4 previous years.

In Cambridge we are keeping a 3D SLIMCAT simulation fairly up to date. Given the cold temperatures it is not surprising that the model pre- dicts strong chlorine activation. Figure 11 shows the model field of ClO_x (= ClO + 2Cl₂O₂) for January 30 1996 at 475 K. The vortex is essentially fully activated, and, even in this low resolution (2.5x5.6), there appears to be some sort of filament leaving the vortex around 180E.

Martyn Chipperfield (Cambridge, ACMSU)

Carbon Aerosols and Atmospheric Photochemistry

Carbon aerosols are produced by all combustion processes. A combustion process of particular importance for our atmosphere is air traffic. A recent paper we have submitted to JGR investigates some possible effects of heterogeneous reduction of atmospheric constituents on carbon aerosols. Reduction of HNO₃ and NO₂ on carbon aerosols may be an important effect of increased air traffic which has not been considered to date. It is shown that if HNO₃ and NO₂ are heterogeneously reduced on atmospheric amorphous carbon aerosols then a significant mid-latitude, lower stratospheric ozone loss mechanism could exist. This ozone loss mechanism is almost independent of temperature and does not require the presence of sunlight. The mechanism can operate at all latitudes where amorphous carbon aerosols are present. The relative importance of the mechanism increases with night-length. The reduction of HNO₃ on carbon aerosols could also be a significant renoxification process wherever carbon aerosols are present. Due to the very different soot levels in the two hemispheres this implies that there should be a hemispheric asymmetry in the role of these mechanisms. The renoxification leads to simulated tropospheric HNO₃/NO_x ratios which are close to those observed. In addition, in contrast to the stratospheric response, the tropospheric production of NO_x due to the reduction of HNO₃ would lead to tropospheric ozone production.

D. J. Lary (Cambridge, ACMSU)

R. Toumi (Imperial)

A. M. Lee (Cambridge, ACMSU)

Chemical Data Assimilation: Ongoing research

The technique of data assimilation is being further developed. The recent SPADE campaign provided a wealth of high time resolution data on a very large number of chemically reactive species. This high time resolution data is being used to critically test the behaviour of the chemical data assimilation and the completeness of our chemical schemes.

At the same time, the technique of chemical data assimilation is being applied to prepare a self consistent data set of atmospheric constituents for a wide variety of chemical constituents and time periods starting with winter of 1991/92. The self consistent set of analyses of chemical constituents could then be used in a wide range of other studies in atmospheric science. This dataset would allow a more accurate investigation of the whole issue of ozone depletion, in particular ozone loss at mid-latitudes which occurs over large population densities such as the United Kingdom. It would also provide insight in to the relative roles of chemistry and dynamics on the behaviour of ozone.

The use of the technique of data assimilation represents an important pioneering first and scientific lead for UGAMP and the United Kingdom. Hopefully it will also allow the United Kingdom to make a substantial contribution to the analysis of EOS data after its launch in the late 1990s. It is hoped that by the next issue of UGAMP news more information will be available.

D. J. Lary (Cambridge, ACMSU)

M. Fisher (ECMWF)

Recent papers

Lary, D. J., Chipperfield, M. P., Pyle, J. A., Norton, W. A., Riishojgaard, L. P., Three dimensional tracer initialisation and general diagnostics using equivalent PV latitude-potential temperature coordinates, 121, 187-210, Quarterly Journal of the Royal Meteorological Society, 1995.

Balluch, M., Lary, D. J., Solar Heating Rates - The Importance Of Spherical Geometry - Comment - Reply, Journal of the Atmospheric Sciences, 52, No.3, pp.383-386, 1995

S. Ghosh, D. Lary And J. A. Pyle, Estimation of heterogeneous reaction rates for stratospheric trace gases with a particular reference to the diffusional uptake of HCl and ClONO2 by polar stratospheric clouds, 13, No. 4, 406-412, Annales Geophysicae, 1995.

Lary, D. J., Chipperfield, M. P., Toumi, R., The impact of the reaction OH+ClO -> HCl+O2 on polar ozone photochemistry, 21, 61-79, Journal of Atmospheric Chemistry, 1995.

Fisher, M., Lary, D. J., Lagrangian four dimensional variational data assimilation of chemical species, Quarterly Journal of the Royal Meteorological Society, Vol. 121, No. 527 Part A, 1,681-1,704, 1995.

Lary, D. J., Gas Phase Atmospheric Bromine Photochemistry, Accepted by the Journal of Geophysical Research, 1995.

Lary, D. J., M.P. Chipperfield, R. Toumi, T.M. Lenton, Heterogeneous Atmospheric Bromine Chemistry, Accepted by the Journal of Geophysical Research, 1995.

Lary, D. J., R. Toumi, A. M. Lee, M. Newchurch, M. Pirre, J. B. Renard, Carbon Aerosols and Atmospheric Photochemistry, Submitted to the Journal of Geophysical Research, 1995.

OH in the night-time troposphere

The OH radical is the dominant oxidant in the troposphere. Until quite recently it was generally believed that the OH concentration decayed to zero at night, however, Tanner and Eisele (1995) have recently reported measurements of OH at night which are about an order of magnitude smaller than those during the day. The big question is how can OH be made at night? During the day the photolysis of ozone produces a small amount of O(¹D), which in turn reacts with water to form OH

O3 + hv -> O2 + O(1D)		(1)

O(1D) + H2O -> OH + OH		(2)

Clearly at night, reactions (1) and (2) cannot operate. An important night-time oxidant is the nitrate radical (NO3), formed from the reaction between O₃ and NO₂,

O3 + NO2 -> NO3 + O2		(3)

During the day NO₃ is rapidly photolysed, but at night concentrations can build up to ca. 20 ppt. Is there a mechanism by which NO₃ can initiate OH production at night? Work carried out in collaboration with Richard Wayne's group at Oxford has shown that a reaction between NO₃ and peroxy radicals (RO₂) can occur. In particular the reaction between the peroxyacetyl radical (CH₃C(O)O₂) and NO₃ has been shown to be fast (Canosa-Mas 1996)

CH3C(O)O2 + NO3 -> CH3C(O)O + NO2 + O2  (4) 

The peroxyacetyl radical (PA) is formed from a number of reactions involving volatile organic compounds during the day, but at night the dom- inant source is the thermal decomposition of Peroxyacetylnitrate (PAN)

CH3C(O)O2NO2 -> CH3C(O)O2 + NO2   (5)

Hence at night the following cycle can occur

CH3C(O)O2NO2  -> CH3C(O)O2 + NO2   (5)

CH3C(O)O2 + NO3 -> CH3C(O)O + NO2 + O2   (4) 

O3 + NO2 -> NO3 + O2               (3) 

CH3C(O)O -> CH3 + CO2                  (6)        

CH3 + O2 + M -> CH3O2 + M       (7)

CH3O2 + NO3 -> CH3O + NO2 + O2      (8)

O3 + NO2 -> NO3 + O2             (3)

CH3O + O2 -> HCHO + HO2          (9)         

HO2 + O3  -> OH + O2 + O2 

-----------------------------------------------------

Net: CH3C(O)O2NO2 + 3O3 
-> HCHO + CO2 + NO2 + 4O2 + OH   

Modelling studies (box) at Cambridge have shown that this cycle could be an important source of night-time OH up to about 4 km over continental regions. Typical night-time OH concentrations range from 1-15% of the daytime value. Globally this may have a significant effect on the trace gas composition of the lower troposphere. Further work is planned, incorporating this cycle into the Cambridge 3-D tropospheric model.

References

D.J.Tanner and F.Eisele, J.Geophys.Res., 100 (1995) 2883

C.E.Canosa-mas, M.D.King, R.Lopez, C.J.Percival, R.P.Wayne, D.E.Shallcross, J.A.Pyle and V.Daele, J.Chem.Soc.Faraday Trans. accepted for publication.

Dudley E Shallcross (Cambridge, ACMSU)

Impact: A new Implicit Timescheme for Atmospheric Chemistry Models

Accurate time integration techniques are essential for 3D atmospheric chemistry modelling, given the fast chemical processing that can occur, for example, on polar stratospheric clouds or across the day/night terminator. At the same time, these schemes must be economical enough to be used in 3D chemical models.

Stott and Harwood (1993) originally proposed an implicit timescheme which achieved good results and was economical through its approximation of the Jacobian matrix. This is a matrix that arises in implicit schemes and couples the ODEs governing the chemical rates of change.

However, it was found that this scheme failed under certain circumstances, more so when the complexity of the chemistry was greater than originally used by Stott and Harwood. Interestingly, some of the problems with the scheme were not noted until fixes in the code designed to prevent negatives were removed (purists take heart!).

Since then work at the ACMSU in conjunction with Peter Stott at Edinburgh has improved the properties of the scheme making it much more stable and robust, particularly when used with chemical families for which the original scheme had problems. The new scheme has been christened IMPACT (IMPlicit Algorithm for Chemi- cal Timestepping). The details of the method are still being finalised and tested but a paper is (finally!) in preparation and hopefully should be submitted before the next UGAMP Newsletter appears.

References

Stott, P.A. and Harwood, R.S., 1993, An implicit time-stepping scheme for chemical species in a global atmospheric circulation model, Ann. Geophysicae, 11, 377-388.

Glenn Carver (Cambridge, ACMSU),

Peter Stott (Edinburgh)

Stratospheric Tracers of Atmospheric Transport

Scientific Overview

The primary goal of the STRAT campaign is the measurement of the morphology of long-lived tracers and dynamical quantities as functions of altitude, latitude, and season in order to help determine rates for global-scale transport and future distributions of high-speed civil transport (HSCT) exhaust emitted into the lower stratosphere. The observations will also improve understanding of broader issues involving transport of gases and aerosols in the stratosphere.

Results from recent airborne campaigns have demonstrated the capability for studies of appropriate combinations of tracers to define atmospheric transport rates for time scales from days to years, and to provide critical tests for atmospheric models used to predict impacts of future fleet emissions.

A secondary goal of STRAT is the further characterisation of atmospheric photochemistry. As shown in earlier airborne campaigns, measurement of free radicals within the context of a sufficiently large suite of tracer observations provides stringent tests of our understanding of the processes that control ozone photochemistry. The STRAT campaign will extend the regions and seasons for which we have such measurements.

To meet these goals, the NASA ER-2 high altitude research aircraft is being deployed to make in situ measurements of chemical species in the lower stratosphere, emphasising extensive sampling from just above and just below the tropopause up to 20 km. The first test flights were conducted from NASA Ames Research Center (37N 122W) in May 1995. Since then the ER-2 has been deployed twice, first in October 1995 and more recently in February 1996. Further deployments are planned for July and October 1996, with further deployments under the aegis of the Photochemistry of Ozone Loss in the Arctic Region in Summer (POLARIS) in May, July and September 1997. During deployment, the following flight scenarios are envisaged:

• Northern survey flight with at least one vertical profile at high latitudes
• Flights to and from Hawaii (and other possible sites)
• Tropical survey flight (from Hawaii) with at least one vertical profile at low latitudes
• Stair-stepping flights from the lowest possible altitude to approximately 70 kft

UGAMP involvement

So far, UGAMP has actively participated in only the February 1996 deployment where its role has centred around forecast support. This support will be provided for future deployments. For preflight planning, we are providing individual flight forecasts, both meteorological and chemical, to help in the identification of optimal mission parameters, including flight location and timing. This type of support is a continuation from that given during SESAME and ASHOE. However, the mission emphasis on dynamics prompts certain modifications to the chemical forecasting methodology. In this situation, the forecasting of fast photochemistry is unwarranted, instead the dynamical state of the atmosphere can be predicted by the passive advection of long-lived chemical tracers at high resolution. High resolution was already recognised during ASHOE as a necessary component of any forecast prepared for an ER-2 type campaign.

In general, the SLIMCAT isentropic chemical transport model advected two tracers: nitrous oxide (N₂O) and potential vorticity (PV). The N₂O was initialised from a low resolution seasonal integration. The 5-day meteorological forecasts were provided by Richard Swinbank and Darren Podd at the UK Meteorological Office. The model grid was optimised for the northern hemisphere where it reached a resolution of 1.0x1.5. The model was run on 10 isentropic surfaces from 330 K to 540 K. For the northern survey flight, a third tracer was included, which was initialised from the low resolution model ClO_x field. This tracer was used to predict air parcels which may had been processed on polar stratospheric clouds within the polar vortex.

The February deployment

The first flight of the deployment on 26 January was a test flight. The next window for a flight given the ground meteorological conditions came on the 29 January. However, the polar vortex was not within range for a northern survey flight out of Ames, so a stair-stepping flight scenario was proposed. The axis of the horizontal flight path was chosen to be perpendicular to the wind jet in the upper troposphere predicted by a meteorological forecast. The northern survey flight did take place on 1 February with the northern terminus chosen to lie above Lake Winnipeg (53N 96W). Both the meteorological forecasts and the chemical forecast predicted penetration of the polar vortex by some 500 km. This prediction was confirmed when preliminary results were presented later in the day, the level of penetration was in fact greater than any flight made during the ASHOE campaign. However, particular interest in this flight arose because of the prominence of filamentary structures associated with recent expulsion from the polar vortex. Figure 12 shows vertical cross sections of forecast modified PV along the proposed flight track (a filament is clearly visible centred at 44.5N and 430 K). Further, the forecasting techniques discussed here indicated that these structures would show the chemical signature of air recently processed on polar stratospheric clouds. The subsequent in situ measurements showed the expected features associated with filamentary structures and the crossing of the vortex edge. Measurements of HCl by the ALIAS infrared spectrometer supported the prediction that this air had been processed recently. The ratio of HO₂ with OH, measured by the Harvard laser induced fluorescence instrument, was low inside the vortex indicating that available inorganic chlorine, Cl_y, had been sequestered largely into the ClO_x chemical family. The Ames deployment was concluded with a second stair-stepping flight on 2 February.

The transit flight to Barber's Point, Hawaii (21N 158W) was delayed for a number of days due to the adverse weather conditions on the ground. A window finally opened on 5 February, but unusually for Hawaii ground conditions were unfavourable for takeoff until 8 February, which was a coordinated flight between the ER-2 and the NASA DC-8 (the DC-8 was taking part in the Tropical Ozone Transport Experiment (TOTE)). The forecast for this flight indicated a high tropopause, which was manifest in the measurements. The southern survey flight took place on 13 February and flew directly south to about 1S. The forecasts for this flight did not indicate anything of interest to divert the flight path from penetrating as deep as possible into the tropical region. The transit flight back to Ames took place on 15 February.

It is doubtful whether forecast winds in the tropics are at all accurate. However, it will be interesting to see whether the tropical structures formed in the model when using winds from a data assimilation system agree with measurements. There are however a number of studies which indicate that the filamentary structures measured outside the polar vortex, but with a polar origin, can be modelled accurately.

Adrian Lee (Cambridge, ACMSU)

© 1996 Centre for Atmospheric Science/UGAMP. All scientific articles are unpublished. No text or graphics may be copied or used without permisson. Newsletter Editor: Glenn Carver, Cambridge University.