Tabulated high resolution cross-sections of O3 measured by Daumont, Brion and Malicet in the early 1990’s [1].
The wavelength range slightly varies with temperature but covers the entire UV to NIR region, from 194.50 nm to
830.00 nm at 0.01 to 0.02 nm resolution. The cross-section data were collected at 0.01-0.02 nm resolution and
each wavelength/cross-section table varies in size from 22,052 to 63,501 entries. The data consists of 5 tables
of wavelength versus cross-section for 5 temperatures.
Notes
Temerature Range
Measurements are provided at 5 temperatures covering typical stratospheric and
tropospheric conditions:
218 K
228 K
243 K
273 K
295 K
Wavelength Range
The wavelength range of each temperature table is slightly different and is given below.
Note that most of the temperature variation occurs in the Huggins band
between 315 and 360 nm:
218K -> 194.50nm to 650.01nm
228K -> 194.50nm to 520.01nm
243K -> 194.50nm to 519.01nm
273K -> 299.50nm to 520.01nm
295K -> 195.00nm to 830.00nm
We looked into temperature interpolation and while DBM suggest that a quadratic interpolation scheme [3] they do
not indicate an explicit technique. We tested several quadratic fitting routines and found that a truncated linear
fit in temperature was visually more appealing than any of the quadratic fits and had none of the undesirable
artifacts (excessive curvature etc.) that naturally arise with quadratic curve fitting. Consequently this object
uses a truncated linear fit in temperature.
Data Source
These data are an exact replication of the data files:
HITRAN UV cross-sections for a given constituent. The constituent must be one of the following:
[“O3”, “NO2”, “BrO”, “SO2”]
These are the files that you obtain from the HITRAN website when selecting UV cross sections.
Note that usually these are not very good, and you should use a specific database for each constituent instead
such as O3DBM. The HITRAN UV cross sections are only provided for convenience.
Specific Issues
O3
We have noticed the Ozone cross sections are unreasonably large in the region < 310 nm, and does not match the source data.
We do not recommend using this.
NO2
Even though the HITRAN database specifically says all cross sections are given in vacuum wavenumbers, this one
seems to be specified in air wavenumbers. We have converted it to vacuum wavenumbers manually. After doing so
it matches the Vandaele source files relatively well.
An optical property that is defined by a database file. This is just a base class to handle file loading,
derived classes must be used as actual optical properties.
Parameters:
db_filepath (Path) – Path to the optical database file
Calculates the Rayleigh scattering cross section in units of [m**2] for a given refractive index according to the Bates
approximation. Cross section is returned back at (0C, 1013.25 hPa)
Parameters:
wavelengths_um (np.array) – Wavelengths in [um]
n2_percentage (float, Optional) – Percentage of N2, default 78.084%
o2_percentage (float, Optional) – Percentage of O2, default 20.946%
ar_percentage (float, Optional) – Percentage of Ar, default 0.934%
co2_percentage (float, Optional) – Percentaage of CO2, default 0.036%
Returns:
np.array – The rayleigh scattering cross section in units [m**2]