GLOBUS campaign of September 1983

00324633/87$3.00+0.00 Pergamon Journals Ltd.

Plartpt. Space Sci.. Vol. 35, No. 5, pp. 66Ml2 Printed in Great Britam.

CONCENTRATIONS OF HYDROGEN CHLORIDE AND HYDROGEN FLUORIDE MEASURED DURING THE MAP/GLOBUS CAMPAIGN OF SEPTEMBER 1983* R. ZANDER, G. ROLAND and L. DELBOUILLE Institute of Astrophysics, University of Liege, Belgium A. J. SAUVAL Royal Observatory of Belgium,

Brussels

P. MARCHl? U.A. CNRS. 776, Reims, France and

F. KARCHER, M.AMOUDEI CNRM, Toulouse, in final

(Received

and B. DUFOUR

France

form11 November 1986)

Abstract-Within the context of the MAP/GLOBUS campaign of September 1983, several trace species have been observed by absorption spectroscopy at the two ground stations of the Jungfraujoch, Switzerland (3580 m altitude) and the Observatoire de Haute-Provence, France (1905 m altitude). The results obtained for HCl and for HF, expressed in terms of mean integrated columns above these sites are :

:

Jungfraujoch

(2.1 k 0.2) El5 mol cm-’ HCl (4.8 f 0.2) El4 mol crne2 HF (2.6 f 0.2) El5 mol cmm2 HCl.

Haute-Provence:

Taking into account the difference in the altitude of the two stations, the reported HCl results are in agreement to within their respective uncertainties. The integrated column density of HCl and HF above 11 km altitude, deduced from airplane observations on 9 September 1983, are : (1.65 + 0.25) El5 mol cm-’ HCl above 11 km (3.7 k 1.7) El4 mol cm-’ HF above 10 km supporting

satisfactorily

the ground

measurements.

INTRODUCIION

Over the last years,

the integrated

HCl

been

and

ground troscopy,

HF

have

through using

studied

near-infrared the

sun

column

densities

of

from

the

intensively absorption

as the

source

sphere. Their relative trends allow the validity of the chemistry being applied in present models to be tested. Furthermore, the HF and HCl measurements, when carried out simultaneously, may help in establishing the relative influence of natural vs anthropogenic sources of atmospheric chlorine. Under ground-based observational conditions, it has been shown that the Rl line of the fundamental band of HCl, at 2925.898 cm-’ and the Rl line of the fundamental of HF, at 4038.965 cm-‘, are best suited for analysis, because of their relative strengths and their isolation with respect to any interfering absorption produced by other telluric constituents, especially water vapor and methane. Airborne observations are obviously less perturbed by these species so that the R2 line of HCl at 2944.914 cm-’ and the RO line of HF at 4000.989 cm-’ can be used as well. In this paper, we report on the results for HCl and HF, deduced from observations carried out during

spec-

of radiation

(Zander et al., 1977 ; Marchit et al., 1980; Marche and Meunier, 1983). While HF measurements are appropriate as an indirect means for evaluating the quantities of chlorofluoromethanes having been destroyed by photolysis in the stratosphere, the concentrations of HCl provide insight into both the total chlorine budget in the atmosphere and the existence of competing reservoirs such as HOC1 and C10N02 in the strato-

*Joint campaign

publication of 1983.

of results

from

the MAP/GLOBUS

665

666

R.

ZANDER et

the MAP/GLOBUS campaign of September 1983 at both the International Scientific Station of the Jungfraujoch, Switzerland (3580 m altitude; 46.55”N; 8.00”E; hereafter Jungfraujoch) by the Liege group, and the Observatoire de Haute Provence, France (Mt Chiran, 1905 m altitude ; 43.88”N ; 6.18”E ; hereafter O.H.P.) by the University of Reims. The results from HF and HCl observations onboard a Caravelle-airplane flying during sunset of 9 September 1983, obtained by the Toulouse group, are also included. That flight was performed at an altitude of 11.9 km, from 45”N. 3”W to 50”N, 1”W; as the sun was observed both above and below the horizontal, the measurements relate to atmospheric layers down to about I1 km.

INSTRUMENTATION

The Jungfraujoch instrument used for the present contribution is a classical grating spectrometer of 7.5 m focal length operating in a double-pass configuration with an intermediary slit; it has been described in detail elsewhere (Delbouille et al., 1973). The spectral resolution attained in the near-infrared is equal to 0.02 cm-‘. Using a liquid nitrogen cooled indium antimonide (InSb) detector, the signal-tonoise ratio is well in excess of 200 for an average of about 50 individual scans recorded at a fast speed of 0.3 cm -’ s-’ ; a limited spectral interval of about 1 absorption line to cm-’ centered on a characteristic be “monitored” can therefore be acquired within a few minutes. The Haute-Provence instrument is a SISAM-type spectrometer. The gratings, 79 lines mm-’ and 220 x 150 mm2, have been used in the seventh order. The spectral region of interest is selected by an appropriate optical filter and a liquid nitrogen cooled InSb detector senses the output radiation.The spectral resolution is equal to 0.02 cm-’ and the signal-to-noise ratio is up to 800 (for small zenith angles), typical spectra (see Fig. 6) are recorded in about one minute. The Caravelle instrument is a “grille” spectrometer coupled to a heliostat which allows infrared solar observations to be carried out during sunset, between 83” and 91.5” zenith angles, through a BaF, window. The instrument has been described by Fontanella et al. (1975). Six spectral intervals of 10 cm-’ each are observed, two of them being assigned to HCl and HF. Each of these intervals is scanned three times consecutively every 5 min; this procedure allows “damaged” spectra to be eliminated and variations of the equivalent widths of absorption lines vs zenith angles to be observed (see Fig. 8). The signal-to-noise

al

ratio is 40 on HCl spectra recorded at a resolution of 0.18 cm-’ whereas it is 10 for HF runs at 0.24 cm-’ resolution. The low S/N ratio for HF is due to the poor response of the PbSnTe detector and optical cut-off near 4000 cm-‘; consequently, only the HF observations at zenith angles larger than 89” have been used in this investigation.

JUNGFRAUJOCH

RESULTS

Hydrogen chloride Only the Rl line of the fundamental band of HCl, located at 2925.898 cm-‘, has been used for the investigation described here. Figure 1 displays a typical spectrum encompassing that HCl line, obtained on 13 September 1983, under a 45.2” zenith angle ; the high frequency noise remaining on this tracing is due to the fact that no filtering has been applied to that sample. Figure 2 represents all the individual values of the equivalent widths (EQWs) vs zenith angle of the HClRl line, measured on the spectra recorded during the various days given in the legend. Within the context of an extended monitoring program, presenting the data in this form avoids the necessity of invoking line parameters which in many instances are not known with satisfactory absolute accuracy. The full curve drawn in Fig. 2 corresponds to the calculated EQWs using : -a multi-layers curved atmosphere; -the mean physical characteristics of the atmosphere prevailing over East France during the campaign ; -the line parameters available in the “1982-AFGL line parameters compilation” (Rothman et al., 1983a, 1983b) with line strengths for HF and HCl adjusted to the recent values of Pine et al. (1985). That profile corresponds to an integrated column of HCl above 3.58 km, equal to 2.11 El5 mol cm-*. Since the full curve of Fig. 2 fits the ensemble of the observational points well, it can be concluded that the mean hydrogen chloride amount above the Jungfraujoch was indeed (2.11 f 0.2) El5 mol cm-* during the MAP/GLOBUS campaign of September 1983. On a day-to-day time scale, the variability in the HCI column was at most + 15% throughout the period covered here. Referring to the work on line intensities by Pine et al. (1985), the estimated absolute uncertainty on the mean HCl column is *9%. (random contributions of 4% for the line strength and halfwidth ; 5% for uncertainty in the continuum setting ; 6% for the limited S/N and 2% for the likely uncertainty in the adopted temperature profile).

Hydrogen chloride and hydrogen fluoride concentrations

667

1.00

H2O 2925.wkm-

HCI- RI 2g25,8g8cm_I

.75

fl

JUNGFRAUJOCH September L5.2°Zenith

13, 1963 Angle

.25

.oo FIG. l.A

I

L

\

I 2927

/

ZERO

I

1 2926

2925 cm-l

TYPICALUNFILTEREDSPECTRUMENCOMPASSINGTHE HCl-RI LINE AT 2925.898 cm-‘, OBTAINED ATTHEJUNGFRAUJOCHON 13 SEPTEMBER 1983, FOR A ZENITHANGLEEQLJALTO 45.2”.

B-l

I

I

‘0

5

I

k 10.0

JUNGFRAUJOCH

E

HCI - Rl Line

I

-MAP-

I

I

I

GLOBUS

v

2925.898cm-1

z 5 w

5.0

1

5

s 2.0

l .

I 30

1.0

I LO

I

50

I 60

I 70

.a

26

.

.a

28

a.

Oct.01

I 80 ZENITH

I 90 ANGLE

”

(‘I

FIG. 2. EQUIVALENTWIDTHSVS

ZENITHANGLEOFTHE HCl-RI LINEMEASUREDON SERIEXOFJUNGFRAUJOCH SPECTRARECORDED DURINGSEPTEMBER 1983. The solid curve corresponds to the EQWs computed with the profile A in Fig. 3.

The daily means of the HCl integrated column above the Jungfraujoch (see Fig. 4) are as follows. Day of Sept. 83 : Molcm-*xElS: Day of Sept. 83 : Mol cn? x E15:

8 2.43 22 1.95

13 2.38 23 1.90

18 2.35 24 1.86

19 2.08 26 1.94

20 2.06 28 2.15

The mean value is 2.11 El 5 mol cm-‘, with a standard deviation of 0.21 El5 mol cme2. Notice that the HCl profile of Fig. 3 is not retrieved from the observations, but a “typical” one fitting the whole EQWs measured for zenith angles between 42” and 88”. During September 1983, the tropopause

668

MAP - GLOWS CAMPAIGN - Sept. 1983 HCL and HF OBSERVATIONS TYPICAL PROFILE COMPATIBLE J’JNGFRAUJOCH EQUIVALENT WIOTHS A=HCL ; C=HF MEAN HCL

E Y.

,

i J

WITH +I9

to50*

10

:

CHIRAN-OHP 20

MEAN HCL AND STD. DEVIATIONS FROM AIRPLANE , BETWEEN INDICATE0 ALTITUDES

-

-iz . z

x 320 tii a

,’ 50 v)

-

:

I 00

-

\

t

E- 12

1

I

rlrl,,

I E- 11

I

CHIRAN-OHP

Irrlll

I

VOLUME

roe

-5

IO0

( 1905m)

1111111

E-10

-2

I

lllll

E-09

E- 08

MJXING RATIO

FIG. 3. TYPICAL VOLUME MIXING RATIO PROFILESCOMPATIBLE WITH HCI AND HF OBSERVATIONS DURING SEPTEMBER~~~~,CARRIEDOUTATTHEJUNGFRAUJOCHANDTI~E O.H.P. STATIONS,ASWELLASFROMONBOARD ACA~~L~AIRPLA~(FORDETAI~,S~~XT~.

height, as measured above Payerne, varied between 9 and 12 km.

Switzerland,

Hydrogenjlttoride

Figure 5 shows all the individual values of the EQWs of the HF-Rl line vs zenith angles, measured from the recordings made during the campaign. The full curve in that figure corresponds to the EQWs computed in the same manner as mentioned for HCl, using the concentration profile represented by C, Fig. 3. The satisfactory agreement between the measurements and calculations indicates that the mean hydrogen fluoride amount above the Jungfraujoch was equal to (4.80 f 0.2) El4 mol cm-’ during September 1983. The variability in the total column of HF among the day-to-day results is f 8%. Based on the HF line parameters published by Fine et al. (1985), the best estimate of the absolute uncertainty is + 7% (4% for line strength and half-width; 3% for errors in the continuum setting ; 4% from the limited S/N and 2% for uncert~nty in the adopted temperature profile). As mentioned before for HCl, the HF profile C of Fig.

3 is a “typical” one, fitting satisfactorily the EQWs measured for zenith angles ranging from 48” to 86”. In Fig. 5, the mean of the HF EQWs at very low sun lies below the computed curve ; this indicates that the tropospheric content of HF is relatively low and does not increase in the lower troposphere as appears to be the case for HCl (see Figs 2 and 3). HF/HCI ratio Based on the results derived for the total column densities of HF and HCl, the mean HF/HCl ratio above the Jungfraujoch was equal to 0.23 _t 0.03 during September 1983. HAUTEPROVENCE

RESULTS

ON HCI

The Rl and PS lines of the fundamental band of HCl, located respectively at 2925.898 and 2775.760 cm-‘, have been recorded, but only RI has been used in the present analysis. Figure 6 shows two spectra obtained on 13 September 1983, for zenith angles equal to 80.70” and 89.95”. On that particular day, a

669

Hydrogen chloride and hydrogen fluoride concentrations

z z -;;

INTEGRATED COLUMN OF HCL MEASURED BY lR-SPECTROMETRY DURING MAP- Globus CAMPAIGN

3.0

I

:

o O.H.P-MT. CHIRAN x JUNGFRAUJOCH

Z

DATE, FIG.

4.

DAILY

MEAN

--I

COLUMN

I

‘0

E

k 10.0 2 5 I=

1983

I

-

I

September

R 1 Line

1993

OF HCl ABOVE THE O.H.P. AND THEJUNGFRAUJOCH, AS MEASURED DURINGSEPTEMBER 1983.

I

JUNGFRAUJOCH HF-

5.0

DENSITIES

I

-MAP

I

I

I

- GLOBUS

L038.965cm-’

-

a I

L

scpt .8

,1993

2

1.0 -

I 30

I LO

I 50

I 60

I 70

. . 26 .* * . Oct. 01 ” I I 90 90 ZENITH ANGLE

(‘1

FIG. 5. EQUIVALENTWIDTHS OF THEHF-RI LINEvs ZENITHANGLE,MEASURED ON SERIES OF JUNGFRAUJOCH SPECTRARECORDED DURINGSEPTEMBER 1983. The solid curve corresponds to the EQWs computed with the profile C in Fig. 3. set of 25 spectra was recorded between 80” and 90” zenith angle ; this enables a “zenith angle curve of growth” to be plotted and information about the vertical concentration profile to be deduced. Figure 7 shows the measured values of the EQWs of the HCl line vs zenith angle. The “zenith angle curve of growth” corresponding to the HCI profile A in Fig. 3 is reproduced as a full line in Fig. 7; it does not fit

the O.H.P. data points, but indicates that the HCl concentration in the troposphere above that station was smaller than that corresponding to curve A, Fig. 3. We have used the method developed by March& and Meunier (1983) to retrieve the most realistic HCl volume mixing ratio profile above O.H.P., adopting an eight-layers model with five layers in the troposphere and three in the stratosphere. The con-

670

R.

ZANDER

et al. The mean value is 2.62 El5 mol cm-‘, dard deviation of 0.10 El 5 mol cm-‘. AIRPLANE

2925.898 cm-’

0.0

0.H.P I”-

OBSERVATIONS 2925.8

2926.0

(Sept. 13,1983)

2925.6

WAVENUMBER

2925.L

I

(cm-‘)

FIG.~. Two O.H.P. SPECTRAIN~~~CINI~OFT~ RI LINE OF HCi, RECORDED ON 13 ~E~MBER 1983, FOR ZENITH ANGLES EQUALTO 80.70” (UPPERTRACE) AND89.95” (LOWER TRACE).

-I

0.H.P

A. Fig.

’

82

.

a4 86 88 ZENITH ANGLE (“1

90

FIG. 7. COVALENT WIDTHS (OPEN D~MONDS) vs ZENITH ANGLE OF THE HCl-RI LINE MEASURED AT O.H.P. ON SEPTEMBER 1983.

The solid curve corresponds to the computed EQWs based on the profile A of Fig. 3 (the EQWs have been normalized to unity for a zenith angle of 0”). ~ntrations were fixed in the three upper layers as the “zenith angle curve of growth” was too noisy to deduce information for the stratospheric region. The retrieved profile is shown as curve B in Fig. 3; this profile was used subsequently to generate synthetic spectra and determine the HCI concentrations best fitting the observations on the other days. The following values have been found for the integrated columns of HCl above O.H.P. (see also Fig. 4). Day of Sept. 83 : Mol cm-’ x El5 : Day of Sept. 83 : M0lcm-~xE15:

Hydrogen chloride Since the spectral resolution (0.18 cm-‘) prevents separating completely the absorption due to the R2 line of HCl at 2944.914 cm-’ from nearby CH, and Hz0 lines, the observed absorption is integrated between 2944.77 cm-’ and 2945.03 cm-’ (interval delineated by vertical dashed lines in Fig. 8). This measured equivalent width (EWm) is calibrated against corresponding integrals (EWc) calculated from synthetic spectra having the same resolution. At each solar elevation 0, the calibration gives a factor Ki by which the reference-HCI vertical profile PO used when computing the synthetic spectrum should be multiplied in order to fit the calculated EWc with the observed EWm (see Fig. 9). If the Ki show no dependence with respect to solar zenith angle, i.e.

CK

o 0.H.P (SISAM)

80

RESULTS

- MAP / GLOBUS

- CURVE

01

with a stan-

4 2.40

11 2.66

5 2.71 12 2.67

7 2.53 13 2.64

8 2.67 14 2.58

9 2.69

then the vertical column density (VCD) of HCl is: VCD (of the measurement) = l? x VCD (of the profile PO). In the general case, a statistical analysis over a large set of possible profiles leading to the independence of the K, vs Bi gives the measured VCD and its limits containing 99% of the solutions. Most likely integrated columns in thick layers are also derived (see Fig. 3). Details of the retrieval method will be published by G. Froment (CNRM, private communication). With the line parameters of the 1982-AFGL compilation and the strength of the HCl line from Pine et al. (1985), the integrated column retrieved for hydrogen chloride is (1.67 + 0.25) El5 mol cm-* above 11 km. The 15% uncertainty is due to an estimated 2% systematic error in the line parameters and 13% random uncertainties (5.5% from the assumed HZ0 concentration and 12% from the EWm variability estimated by the retrieval procedure). Notice that the beginning of the flight was stratospheric up to 48.5” North and that the VCD of H,O changed regularly by a factor of 0.5 from 45” to 50” North. This evolution of HZ0 has been deduced from two spectral intervals near 1325 and 1603 cm-’ and used in the synthetic spectra calculations. Hydrogen fluoride HF measurements were performed under the same observational conditions as HCl, but their S/N ratio

Hydrogen

0

I

chloride

and hydrogen

I

I

I

671

fluoride concentrations

I

I

I 2940

2946 WAVENUMBER (cm-‘)

FIG. 8. Two SETS OF THREESPECTRA(SOLID CURVES)BETWEEN2939.5 AND 2946.5 cm-‘, RECORDEDWITH THE“GRILLE” SPECTROMETER ON BOARDTHECARAVELLEAIRPLANEAT 11.9 km ALTITUDE.

The dotted tracings represent the computed absorption characteristics for observational zenith angles of 86.3” (upper curve) and 90.5” (lower curve). The vertical dashed lines show the limits adopted for the EQWs measurements of the HCl-R2 line at 2944.914 cm-‘.

is

much lower for the reasons mentioned

venting file”.

therefore

With

of Fig.

a profile

3, the

was found

before,

the retrieval of a “most analogous

vertical

to be equal

column

to the distribution density

pre-

likely pro-

above

C

10 km

to (3.7 Ifr 1.79) El4 mol cmm2.

CONCLUSIONS

Comparison of the hydrogen chloride results obtained during September 1983 at the Jungfraujoch and at O.H.P. indicates a reasonably good agreement for the total column content of HCl. Considering the difference in altitude between the two stations, the integration of profile B of Fig. 3, between 1905 and 3580 m, corresponds to about 0.1 El5 mol cm-‘, which is within the experimental errors of both measurements. Comparing only the results on those days when observations were made at both sites leads to an even better agreement (see also Fig. 4). The airplane-measured column of HCl above 11 km is consistent with the ground-based measured monthly means, if the integrated column between 3.58 and 11 km is less than 9.0 El4 mol cm-’ (the profile A of Fig. 3, integrated between these altitudes gives 1.8 El4 mol cm-‘). The airplane-retrieved profile of 9 September

1983 suggests that the bulk of HCl extended a few kilometers lower than the “typical” mean profiles fitting the September ground observations (curves A and B in Fig. 3). It is worth noticing that the largest column densities above the Jungfraujoch (approx. same latitude as for the airplane’s measurements) were observed during the first half of September ; this may well have reflected a temporary enrichment of HCl in the lower stratosphere, compatible in magnitude with the airplane findings. The mean airplane-retrieved HF column density observed above 10 km is in agreement with the much more reliable column density observed above the Jungfraujoch station. The Jungfraujoch and O.H.P. data sets reported here are the results of monitoring programs intended to evaluate the secular increase of HCl in the atmosphere. On a statistical basis, the results gathered to date at both stations (since 1977 at the Jungfraujoch ; since 1978 in Reims then at O.H.P.) show no significant increase of the integrated column of HCl, in contradiction with findings of Mankin and Coffey (1983). The variability frequently observed in the ground measurements appears to be due, in large part, to tropospheric HCl which renders trend retrievals

R. ZANDEX et al.

672

‘0

50 ‘; 8, ,, E g s

VCD =

2.12

x

lo

30

E ti s g El

10

I

I

90

85 ZENITH

ANGLE

92 (")

FIG. 9.CURVE OF GROWTH OF THE EQWs OF m HCl-R2 LINE vs ZENITH ANGLE MEASUREDFROMONBOARD THE~ARAVELLEAIRPLANE (CROSSES). The open squares correspond to EQWs calculations using the smoothed four-stairs profile delineated by

arrows in Fig. 3; the open triangles are for similar calculations with that same profile multiplied by 1.3 (the vertical column densities, VCD, for these two profiles are given in the figure). The full curves are polynomials of the fifth order fitting each set of points. somewhat more difficult, requiring an extended time base for a firm assessment of that important matter. The HF results deduced from the September 1983

observations at the Jungfraujoch are in good agreement with previous and subsequent findings. Based on data collected since 1976, the yearly increase of the total column of hydrogen fluoride above that station is equal to (10 f l)%. HF columns reported by Marchi et al. (1980), by March6 and Meunier (1983) and by Mankin and Coffey (1983) support such an increase. Notice that an extensive discussion on the monitoring program of HCl and HF at the Jungfraujoch station and related findings (trends and variability) are given in papers by Zander et al. (1987a, b).

derived from stratospheric Appl. Optics 14, 825.

absorption

infrared

spectra.

Mankin, W. G . and Coffey, M. T. (1983) Latitudinal distribution and temporal changes of stratospheric HCl and HF. J. geophys. Res. 88, 176. March& P., Barbe, A., Secroun, C., Corr, J. and Jouve, P. (1980) Mesure des acides fluorhydrique et chlorhydrique dans l’atmosph&e par spectroscopic infrarouge g partir du sol. C.R. hebd. Skanc. Acad. Sci. Paris 290,369. March&, P. and Meunier, C. (1983) Atmospheric trace species measures above Haute-Provence Observatory. Planet. Space Sci. 31, 13 1. Pine, A. S., Fried, A. and Elkins, J. M. (1985) Spectral intensities in the fundamental band of HF and HCl. J. molec. Spectrosc. 109, 30. Rothman, L. S., Gamache, R. R., Barbe, A., Goldman, A., Gillis, J. R., Brown, L. R., Toth, R. A., Flaud, J. M. and Camy-Peyret, C. (1983a) AFGL-atmospheric absorption line parameters compilation : 1982 Edition. Appl. Optics 22, 2247.

Acknowledgements-The results from the Jungfraujoch, whose operation is supported by the Belgium FRFC, have been obtained within the frame of a sustained effort initiated by the Chemical Manufacturers Association, Washington, DC. We thank J. Bosseloirs and J. Deprez for their help in assembling this paper.

REFERENCm Delbouille, L., Roland, G. and Neven, L. (1973) Photometric atlas of the solar spectrum from 3000 to 10000. Special Vol., Institut d’Astrophysique de l’Universit.5 de Litge. Fontanella, J. C., Girard, A., Gramout, L. and Louisnard, N. (1975) Vertical distribution of NO, NO2 and HNOp as

Rothman, L. S., Goldman, A., Gillis, J. R., Gamache, R. R., Pickett, H. M., Poynter, R. L., Husson, N. and Chedin, A. (1983b) AFGL-Trace gas compilation: 1982 version. Appl. Optics 22, 1616. Zander, R., Roland, G. and Delbouille, L. (1977) Confirming the presence of hydrofluoric acid in the upper stratosphere. Geophys. Res. Lett. 4, 117. Zander, R., Roland, G., Delbouille, L., Sauval, A., Farmer, C. B. and Norton. R. H. (1987a) Monitoring of the integrated column of hydrogen fluoride above the Jungfraujoch station since 1977-the HF/HCl column ratio. To appear in J. atrnos. Chem.

Zander, R., Roland, G., Delbouille, L., Sauval, A., Farmer, C. B. and Norton, R. H. (1987b) Column abundance and long-term trend of hydrogen chloride (HCl) above the Jungfraujoch station. To appear in J. atmos. Chem.