Kandis Lea Jessup and Mark Bullock

1
DETAILED CALCULATIONS OF THE VENUS SPECTRUM FROM
0.8-2.5 mm

• Kandis Lea Jessup and Mark Bullock
• Southwest Research Institute (SwRI)

Results
Model Inputs
Model Methodology Radiative Transfer models
provide the basic means for analyzing
near-infrared spectra of Venus nightside and
deriving estimates of gas abundances and cloud
properties. We use the Toon et al. rapid
two-stream hemispheric mean approximation to the
radiative transfer equation to model the expected
thermal emission from the top of the atmosphere
(TOA) in the 1.2 and 2.3 micron regions. This
method includes multiple scattering in infrared
bands for flux calculations as well as the use of
the Planck source function for emission in the
infrared. The hemispheric mean approximates the
angular dependence of the intensity so that the
flux and intensity from an isotropic source are
correctly related.   The temperature T,
pressure, P, profiles and gas density profiles
for the RT calculation are derived from the Venus
International Reference Atmosphere (VIRA). The
gas opacity within the Venus atmosphere is
assumed to be a function of the contributions
from CO2, H2O, SO2, CO, HDO, OCS, H2S, HCL and
HF, where the mixing ratio for each of these
species as a function of altitude is given in
Figure 3. The atmospheric water vapor mixing
ratio appears to be near constant below the Venus
clouds, at about 30 ppmv. It decreases within
the clouds as it is taken up in liquid H2SO4
aerosols, and achieves an abundance of 1 ppmv
above the clouds. Variation in these species is
likely, e.g., CO has been seen to generally
increase with altitude just below the clouds.
Since it is oxidized by the products of the
thermal decomposition of H2SO4, its abundance
may vary spatially and temporally as well.   Also
included in the total atmospheric opacity are
(van de Hulst, 1981) Rayleigh scattering due to
CO2 and N2 and the CO2 (Moskalenko et al. 1979)
and H2O (Liou, 1992) continuum opacities (see
Figure 4). As Figure x indicates, the CO2
continuum opacity is poorly known. For simplicity
we did not include Mie scattering of the H2SO4
particles. Although particle scattering does
strongly affect the overall flux levels, it has
no effect on the bands associated with the trace
species.   Figure 5 shows calculations of the
thermal emission expected at the TOA based on
integration of the flux per atmospheric level for
the nominal Venus atmosphere model depicted in
Figure 4. In Figure 6 the emission expected at
the TOA based on varying the CO and H2O levels to
½ and 2 x the nominal values indicated by the
VIRA data based on the integration of the flux
over the altitude region extending from 100 down
to 20 km. Thus focusing on the altitudes where
the 2.3 micron radiation due to CO originates and
the 1.0-1.35 micron radiation due to H2O is
dominant.   For these calculations the gas
opacity is determined from the correlated k gas
absorption coefficients derived from detailed
line-by-line (LBL) calculations from optical data
obtained from the HITRAN and HITEMP databases
assuming the range of temperatures, pressures,
and gas mixing ratios indicated by VIRA (see
Figure 7). Individual spectral lines, modeled by
rapid Voigt profile calculations, are resolved
with approximately 10 wavenumber points in their
cores. Spectral line strengths in the wings are
determined to about 6 half-widths on either side
of line center because of the importance of
extending the wing calculations much further from
spectral line centers (up to 120 half-widths).
Additionally, the effects of line mixing is
accounted for. Line mixing is significant for CO2
and other molecules. In the current model, line
mixing is handled by using the ab initio methods
of Hartmann and Boulet (1991) to account for line
mixing effects in CO2 at 10 bars and 400 K. For
H2O, the finite duration of molecular collision
results in line profiles that are
super-Lorentzian. This is accounted for by the
use of simple empirical coefficients, c, so the
wider shape is accounted for, significantly
increasing the contribution from the wings of the
absorption lines. Details of the c-factors for
H2O are given by Clough et al. (1989). See Goody
and Yung (1989) for a similar treatment for
CO.   The k-distribution concept postulates that
if the LBL calculated absorption coefficients are
redistributed based on their strength, the
absorption coefficients can be subdivided into a
suitable number of bins arranged as a smooth
function of the absorption strength, known as the
k-distribution according to a new frequency
variable, g. The essence of the ckd scheme is
that this smooth function can be subdivided into
a smaller number of bins according to the rate at
which the absorption strength varies as a
function of the frequency space variable, g.
These same bins can be used at various
temperature and pressures. That is, as long as
the line center remains fixed, independent of the
pressure or temperature, then the rank of the
absorption coefficient strength and the
wavelength remains fixed as a function of
pressure and temperature. Using the techniques
described in Lacis and Oinas (1991) we have
developed both a scheme for producing the
k-distribution of the LBL produced absorption
coefficients and for reducing the k-distribution
into bins of length Dgn, where n is the
subinterval number. For these calculations we
split the 0.8-2.5 micron region into 67 spectral
intervals with an average spectral width of 0.02
microns, and then split each of these spectral
intervals into 16 subintervals or gauss
points. When combined with rapid two-stream flux
calculations, the ckd method is ideal for
high-accuracy General Circulation Models (GCMs)
and automated rapid remote detection of dangerous
gases. The effects of multiple scattering can be
included if single scattering albedos, scattering
asymmetry parameters, and abundance of
atmospheric aerosols are estimated or provided.
The much more commonly used spectral band models
(Goody and Yung 1989) cannot be integrated with
multiple scattering calculations, making them
inaccurate for cloudy (Earth and Venus) or dusty
(Mars) atmospheres.  
Abstract Retrievals of trace gas abundances deep
in Venus atmosphere are best achieved by
observing the nightside of Venus between 0.8 and
2.5 microns. At high spectral resolution, it is
possible to obtain H2O, CO, SO2, CO, OCS, HCl and
HF abundances. Accurate synthetic spectra are
necessary to assess how the emission spectrum in
this region is affected by abundances of these
gases. We have developed a very detailed
line-by-line computer code that accounts for the
non-Lorentzian nature of CO2 and H2O lines far
from line centers in this region, and line mixing
within the P, Q, and R branches of the 15 micron
CO2 band. In addition, we have developed a highly
adjustable correlated-k scheme for calculating
absorption coefficients. We will show synthetic
spectra of the Venus atmosphere in emission form
0.8 to 2.5 microns, and demonstrate the
sensitivity of outgoing flux to changes in trace
gas abundances. Demonstrating that changes in the
atmospheric emission can be quickly and
accurately determined from opacities calculated
from absorption coefficients derived from the
correlated k scheme
INSERT L 1-81 RADIANCE VS. POLLACK CO2 FIGURE
Figure 3
Venus Atmosphere Observations
Venus 2.3 mm IRTF Observations acquired by
Elliot Young and Mark Bullock
INSERT L 1-60 VARIABLE SPECIES RADIANCE FIGURE
a
Figure 4
May 4, 2004
May 5, 2004
May 6, 2004
May 8, 2004
May 9, 2004
May 10, 2004
Figure 1 Images of Venus nightside at 2.3 mm,
taken with the SpeX instrument at the IRTF on
Mauna Kea. SpeX acquires images and spectra
simultaneously (the dark line is the spectrometer
slit, 0.3 x 60). By allowing the slit to drift
across Venus over about 10 minutes, image cubes
from 0.8 to 2.5 mm were obtained. Thermal
radiation orginating from 30-35 km silhouettes
Venus lower clouds, so that where the image is
dark, the lower clouds are thickest. Venus
appearance changes dramatically from one morning
to the next, due to the 6 day rotation of the
cloud deck at this level.
b
INSERT MBS LBL FLUX FIGURE
INSERT LBL ABS COEFF. EXAMPLE FIGURE
INSERT MARQ 2.3 SPECTRUM FIGURE
Conclusions Gaseous opacities derived using the
correlated k scheme leads to in efficient and
accurate calculations of the changes in thermal
emission within a dense atmosphere like Venus due
to changes in the abundance of trace gas species.