Middle Atmosphere News from Oxford

Warwick Norton continues to explore the behaviour of the middle atmosphere in the EUGCM, and compare it with the observed middle atmosphere. Among other things he and John Thuburn have found a '2-day wave' in the model; this is a mainly mesospheric phenomenon that has been known observationally for many years. The EUGCM simulation is apparently the first good representation of the 2-day wave in a GCM. A paper has been submitted to GRL on this work. Suzanne Rosier has finished a series of experiments with the UKMO Stratosphere-Mesosphere Model, for comparison with the 1991-92 northern winter as observed by ISAMS. She is currently completing the writing-up of her DPhil thesis. Helen Rogers continues to carry out experiments with the SLIMCAT model and a Contour Advection model to simulate the transport of sulphuric acid (associated with Pinatubo aerosol) in the lower stratosphere. She has used both UKMO assimilated winds and ECMWF analysed winds to drive the transport models, and finds interesting differences in low latitudes between the two sets of runs. Euain Drysdale is proceeding with his use of the UKMO SMM to model equatorial waves in the lower stratosphere. Alan Iwi has been performing contour-advection calculations to simulate ISAMS observations of N₂O structures in low latitudes.

David Andrews (Oxford)

Report from Imperical College of Science, Technology and Medicine

Science research

(a) Wenyi Zhong

i) Stratospheric infrared heating rates due to volcanic sulphur dioxide.

Although the largest immediate radiative impact of a volcanic eruption is due to dust and sulphate aerosol it is possible (Slimane Bekki) that high concentrations of sulphur dioxide gas may persist in the stratosphere months or years after a massive eruption. Previous studies (Lary et al. 1994, Gerstell et al. 1995) have shown the importance of SO₂ ultraviolet absorption to stratospheric heating rates but its infrared effects have generally been considered to be small. By considering the three infrared absorption bands of SO2 we have shown that, depending on the vertical extent of the SO2 plume, the infrared effects may, in fact, be of similar magnitude to, or larger than, the solar heating, enhancing it in the lower stratosphere and acting in opposition at higher altitudes. This heat source/sink may have significant impact on middle atmosphere temperature, chemistry and circulation following a massive volcanic eruption. A paper has been submitted to QJ.

ii) Radiative forcing of climate by stratospheric and tropospheric ozone.

Various assumptions concerning actual and potential changes in ozone concentration have been used both in Ôoff-lineÕ radiation code and within the 2-D model to investigate their impact on radiative forcing at the tropopause. Initial results suggest a significant seasonal variation such that received wisdom (e.g. that strat O3 reduction tends to result in negative radiative forcing) is only true as a gross global average.

(b) Ian Russell

Antarctic polar vortex response to a prescribed ozone hole.

This research is concerned with the effect of changes in polar ozone on the radiation balance, temperature structure and evolution of the polar vortex in the winter high latitude stratosphere. The model used is the 19 level T42 UGCM2.1 with seasonal cycle and external ozone climatology.

The new diagnostic update (written with Conrad) is used to output diagnostics such as long-wave/short-wave fluxes and heating rates. Daily average radiation diagnostics are output every two days both as zonal means and as lat.-long. plots at different levels.

Three runs have been carried out with initial prescribed ozone depletions of 25%, 50% and 75% between 60S and 90S and between 20mb and 200mb during the period 20 August to 20 November.

Preliminary results show that the ozone depletion causes a delay in the final warming and vortex breakdown of up to 30 days. This is the subject of further analysis.

(c) Simon Ibbotson

Modelling the extratropical Quasi-Biennial Os- cillation.

Studies of the interannual variability in the extratropical QBO have been limited in the past due to restrictions in computing time involved in three dimensional modelling. Research has been undertaken here using the two-dimensional dynamical, chemical, radiative interactive model of J.Kinnersley with three parametrized waves to model the changes in the wave-mean flow interactions associated with the QBO over a decadal timescale. The aim of this research is to determine the causes of the observed 10-12 year temperature oscillation in the northern polar winter stratosphere. Preliminary results show a 10-12 year temperature oscillation consistent with observational measurements.

(d) Jo Haigh

Computing

Conrad Winchester

i) Speeding up (E)UGCM longwave radiation code (with Wenyi Zhong)

The look-up-table approach for CO₂ and O₃ transmittances is simple and very accurate (because it uses line-by-line calculations including Doppler broadening and two-parameter scaling approximation) but is fairly slow in implementation in the 47 level model. We are now fitting an analytic function to the CO₂ and O₃ transmittances over a range of pressures and absorber amounts that are wide enough so that each function can be used over the full 47 levels. By fitting first in pressure space we achieve a general accuracy to 5 decimal places. We are now incorporating temperature dependence. The new scheme should considerably enhance speed without significant loss of accuracy.

ii) The libFutf library is now implemented in IDL.

It is quicker, more memory dynamic and uses structures to store data about UTFs. This means that UTF fields can now be easily read into IDL and then manipulated by it. It is fully documented but needs further testing.

iii) Time series analyser of UTFs.

Using the above library a full blown interactive time series analyser of UTFs has been written. A user specifies a full UTF file that is scanned to find out what fields are in it. These fields are then listed and the user chooses what variable is required for analysis.

The program then extracts all the occurrences of that variable in the UTF noting on what days the variable is defined and presents the user with a popup window showing the variable (It can be either a Lat-Long or Lat-Pressure/height plot), on the day chosen, and a menu of functions to perform. Using a mouse or slider the user can specify a point on the plot from where the function is performed and then call up the function. The present functions supported are

a) 1-D time series - the time variation of the variable at the point specified.

b) 2-D horizontal time series - The time variation of a horizontal slice through the point specified.

c) 2-D vertical time series - The time variation of a vertical slice through the point specified.

d) Vertical slice (either pressure-v-variable or lat-v-variable) e) Horizontal slice (either Lat-v-variable or long-v-variable)

f) FFT - fourier transform of the variable in both 1-D and 2-D. N.B. In the above vertical means the vertical variable on the plot and so for a lat-long plot this would be latitude and for a pres- sure-lat plot it would be pressure. The same goes for the horizontal variable as well which is longitude or latitude respectively.

The program works with UTFs produced from UMAP and the radiation diagnostic routines developed here at Imperial.

This program is continually being developed and we would welcome suggestions for more applications.

If you want information on any of the above, or would like to try out any of the software I can be contacted now by mail to c.winchester@ic.ac.uk

Publications

Toumi R, J D Haigh and K S Law "A tropospheric ozone - lightning feedback" accepted by GRL.

Zhong W, J D Haigh, R Toumi and S Bekki "Infrared heating rates in the stratosphere due to volcanic sulphur dioxide" submitted to QJRMS.

Jo Haigh (Imperial College)

Edinburgh News

High Resolution studies of Rossby Waves using the USMM

Gordon Watson and Peter Stott have been doing high resolution studies of Rossby wave breaking in the USMM on the C90 at Cineca. Initial results indicate that the mechanism of vortex erosion and eventual break-up is much more rapid at T106 than at T42. The model develops a pronounced surf zone with small and irregular PV gradients separated from mid-latitudes by a sharp PV barrier. Streamers are seen to be pulled off the edge of the vortex and mix into the surf zone more easily, leaving a smoother vortex with much sharper PV gradients at the edge than at T42.

So far it has only been possible to run the model for 23 days with tropospheric boundary forcing during a warming period in early February 1992. During this period the atmosphere develops a wave 2 type minor warming with a clear tropospheric precursor. The warming is reproduced in the model at T42, but the vortex doesnÕt split. At T106 the vortex is observed to split more easily and erode much faster than at T42 (see figure 13) with some gross differences towards the end of the run. During this warming period, increases in both vertical and horizontal resolution can alter the ease with which the vortex splits and we are investigating whether this behaviour is consistent with a similar wave 2 warming period in 94/95. Gordon has also been doing high resolution integrations with tracers during a much smaller warming event in December 94 to see how much faster vortex erosion is at T106 with smaller wave forcings.

If anybody has a spare 5x108 Mflops (or about 500 hours of YMP time), we think it would be interesting to do a full annual integration of the model at T106!

Gordon Watson & Peter Stott (Edinburgh)

Rutherford Appleton Laboratory News

For a variety of reasons, the UGAMP group at RAL has and is suffering a bit of an exodus as people depart for pastures of a different hue: Barbara Cox left us in November to take up the pleasures of life in the Fenlands, Sarah Ruth in February for a permanent position in NERC administration and Bryan Lawrence will be leaving in August for a lectureship in New Zealand.

Meanwhile, work carries on.

TOASTE-C

HALOE

Lesley has been looking at some of the HALOE water vapour distributions, particularly looking at the slopes of the isolines in the subtropics for evidence of interannual variations. Although the work is preliminary, the HALOE data appear to have a stronger interannual signal than previous studies (using, for example, the SAGE data).

Together with Lesley and Sarah, Richard Kennaugh has recently completed a comparison of the interannual variability in the double peak seen in HALOE methane and in an isentropic 2-d model. This work is currently being prepared for publication.

QBO

Currently underway is the study of long-term (decadal) variation in the gridpoint SMM, in particular we are looking for systematic variations when a realistic QBO is forced in this model.

Since submitting a paper with Bill Randel on re-analysis of PMR data, Bryan has been buried in the gridpoint SMM model. He's extended the physics to include dry convection and more options for gravity wave drag, including both the HineÕs scheme and the FrittÕs scheme. HeÕs now doing simulations to look at the impact of these schemes on both the Eulerian and Lagrangian circulations. HeÕs also been redoing and reanalyzing a series of seasonal integrations aimed at examining the internal stratospheric transience.

Lesley Gray (RAL)

© 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.