            ClNO3 PSEUDO-LINELIST

INTRODUCTION.
This documents the ClNO3 pseudo-linelist derived at JPL
in November 2000 from the laboratory spectra of Manfred
Birk and Georg Wagner, whose work was described at the
2000 HITRAN conference. The measurement conditions for
each of these spectra are tabulated below. Each of these
spectra covers the 690 to 1330 cm-1 region.

 #  P_tot  P_clno3  Temp    Resn  Noise  File
-----------------------------------------------------
 1  1.068  1.0680  297.02  .0020  0.000  "missing   "
 2  0.558  0.5577  297.20  .0020  0.497  "297_000.xy"
 3  95.08  1.0161  297.39  .0020  0.198  "297_100.xy"
 4  44.76  1.0065  297.44  .0040  0.331  "297_050.xy"
 5  23.91  1.0121  297.13  .0020  0.305  "297_025.xy"
 6  1.013  1.0131  297.00  .00094 0.000  "missing   "
 7  0.321  0.3206  189.58  .0020  0.433  "190_000.xy"
 8  96.24  0.6041  189.58  .0055  0.000  "missing   "
 9  94.59  0.3027  189.58  .0055  0.573  "190_100.xy"
10  44.96  0.3015  189.58  .0040  0.382  "190_050.xy"
11  24.57  0.3061  189.58  .0020  0.542  "190_025.xy"
12  75.73  0.3017  189.70  .0055  1.000  "190_075.xy"
13  12.74  0.2999  189.70  .0020  0.717  "190_012.xy"
14  155.8  0.3022  189.70  .0055  0.480  "190_150.xy"
15  0.603  0.6028  189.60  .00094 0.000  "missing   "
16  95.42  0.6209  219.24  .0055  0.285  "219_100.xy"
17  0.303  0.3026  219.00  .0020  0.761  "219_000.xy"
18  154.2  0.3044  218.93  .0083  0.924  "219_150.xy"
19  53.40  0.3058  218.95  .0040  1.504  "219_050.xy"
20  26.10  0.3061  218.94  .0020  0.940  "219_025.xy"
21  12.83  0.3033  218.95  .0020  1.060  "219_012.xy"
22  5.225  0.3007  218.95  .0020  0.752  "219_005.xy"
23  147.8  0.3029  249.25  .0083  0.516  "249_150.xy"
24  98.30  0.6034  249.24  .0055  0.425  "249_100.xy"
25  25.59  0.3135  249.23  .0020  1.590  "249_025.xy"
26  53.53  0.6078  249.24  .0040  0.346  "249_050.xy"
27  12.72  0.3032  248.41  .0020  1.295  "249_012.xy"
28  0.303  0.3025  248.27  .0020  1.607  "249_000.xy"
29  13.86  0.3028  204.42  .0020  1.000  "204_012.xy"
-----------------------------------------------------
P_tot is the total pressure in mbar
P_clno3 is the ClNO3 partial pressure in mbar

DESCRIPTION.
First, the cross-sections were converted back into transmittance
spectra from knowledge of the cell length and gas concentrations.
In transmittance, the noise will be much more constant than in
absorptance, especially considering that in some of the spectra,
the depth of the 780 cm-1 Q-branch is over 80%. The resulting 25
laboratory transmittance spectra were then simultaneously fitted
(using the GFIT algorithm) by iteratively adjusting the strengths
and ground-state energies of the pseudo-lines.

PSEUDO-LINE SPACING.
After trying various line spacings, an interval of 0.01 cm-1 was
finally chosen.  This difficult choice was a trade-off between:
   1) minimizing the number of ClNO3 pseudo lines,
   2) adequately representing the ClNO3 spectral structure,
   3) getting the correct slant column abundances of ClNO3.

Having more pseudolines (e.g. 0.002 cm-1 spacing) provides better
representation of the spectral structure observed in the low pressure
laboratory spectra, and gives slightly more accurate representation of
their slant column abundances, but the differences were small ( < 1% ).

But most of the atmospheric ClNO3 resides at pressures (25-75 mbar)
where this fine structure is smeared by pressure-broadening. So
in terms of correctly representing the ClNO3 absorption in spectra
of the Earth's atmosphere, 0.01 cm-1 spacing seems to be adequate.

DISCUSSION:
Although the new laboratory ClNO3 measurements of Birk and Wagner
are clearly the best to date, they are not perfect. Some minor
deficiencies became apparent during the fitting of the laboratory
spectra, that users of the raw cross-sections should be aware of:
1) Channel fringes of up to +/- 1% amplitude.
2) HNO3 absorptions of up to 2% depth are noticable in the 1310-
   1330 cm-1 region of spectrum #2 (297K; no air broadening).
3) H2O absorption lines are noticable in the 1312-1321 cm-1
   region of most spectra. # 16 (219K and 95 mbar) is especially
   prominant; the 1318.929 cm-1 H2O line attaining 8% in depth.
   Note that these H2O lines are formed at room temperature and
   at low pressure (inside the FTIR ?), not in the cell containing
   the ClNO3.
To minimize propagation of these artifacts into the ClNO3 pseudo-
linelist, channel fringes, H2O, and HNO3 were all fitted during
the pseudo-line iteration process. Otherwise, when analyzing
atmospheric spectra in the 1320 cm-1 region, H2O and HNO3
absorptions would be misidentified as ClNO3, and vice versa.

IMPLEMENTATION.
To make best use of these pseudo-lines, the assumed ClNO3
Doppler width must be approximately equal to the 0.01 cm-1
pseudo-line spacing. Otherwise the ClNO3 spectrum will start
to appear "spikey" in the upper stratosphere. If your analysis
software calculates the Doppler width based on the molecular
weight of ClNO3, I suggest reducing it from 97 to 1.
   Of course, this problem could have been completely avoided by
having the ClNO3 pseudo-lines on a 0.001 grid, but then there would
have been 40 Mbyte ClNO3 pseudo-linelist, rather than just 4 Mbyte.

CALCULATION OF S AND E"
At each line frequency, an effective strength and ground-state energy
was derived by non-linear least squares fitting to the 25 spectra.
Prior to this a PBHW of 0.078 cm-1/atm had been chosen since it gave
the best overall fits to the pressure dependence of the 780 cm-1
Q-branch region. As part of the fitting, the strengths and ground-state
energies were both constrained to be +ve.

After deriving the pseudo-line strengths, we noticed that in certain
spectral regions ( 880-965, 1005-1090 and 1130-1215 cm-1) the
strengths were all very small. We therefore deleted all the lines
from these intervals to minimize the size of the pseudo-linelist
and to save time when performing line-by-line calculations in those
spectral regions.

Since the Birk laboratory spectra stop at 1330 cm-1, the new ClNO3
pseudo-linelist still contains the old pseudo-lines for the 1680-
1790 cm-1 region, which were based on the measurents of Ballard et al.
(1988).  Note that the 1330-1340 cm-1 region, which was previously
covered by Ballard's cross-section, is no longer represented in the
new list.

The new ClNO3 pseudo-linelist therefore contains lines in the
following 5 intervals: 

       V_start  V_stop     N_lines  Delta_nu   Spectra
        690  -  880 cm-1    19000    0.01    Birk & Wagner
        965  - 1005 cm-1     4000    0.01    Birk & Wagner
       1090  - 1130 cm-1     4000    0.01    Birk & Wagner
       1215  - 1330 cm-1    11500    0.01    Birk & Wagner
       1680  - 1790 cm-1     1540    0.07142 Ballard et al.

Comparisons of ClNO3 retrievals from MkIV balloon spectra in the 780 cm-1
region, suggest that the new pseudo-linelist produces 5-10% smaller ClNO3
amounts than either the Bell linelist or the previous pseudo-linelist
derived from the Ballard cross-sections. 
Furthermore, the new cross-sections of Birk and Wagner provide much
better consistency between ClNO3 amounts derived from the 780 and
1290 cm-1 regions (previously the latter gave 15% smaller amounts).
----------------------------------------------------------------------
