First Steps Towards Putting Earth's Atmosphere in My Model

There are dozens of small steps that have to be taken to make incremental progress on a research project. This was one of those steps that ended up providing a huge boost in motivation and satisfaction.

The image to the left is figure 2.12 from "Fundamentals of Atmospheric Radiation: An Introduction with 400 Problems" (2006, Wiley-VCH) by Craig F. Bohren and Eugene E. Clothiaux. It shows the radiative quantity absorption cross section in units of square micrometers for several gaseous constituents of earth's atmosphere. Square micrometers is a unit of area, but this quanitity doesn't tell us the physical cross sectional area of the molecule, but rather a representative cross sectional area that depends on the wavelength of light it's absorbing. Note the range on the x-axis extends from wavelengths of 1 micrometer (near the ultra violet portion of the spectrum) to wavelengths longer than 1 milimeter.

I've used pink lines to highlight the range that corresponds to the figure on the right. This figure was created by me from the absorption cross sections I retrieved from the HITRAN 2012 database using the radiative transfer model ARTS. Although the horizontal scale is stretched you can see that there is excellent agreement between the two figures. Most of the major features are reproduced. In green circles I've highlighted the few regions in which major features arenot matching up.

The first is in the visible portion of the spectrum for ozone. This is called the Chappuis band. The HITRAN database does not contain spectral line information or absorption crossections in this portion of the spectrum for ozone and since ARTS is more typically used at longer wavelengths it does not yet contain any continuum absorption models to reproduce it.

The second missing feature is in the infrared portion of the spectrum for oxygen. Absorption here is caused by molecular collisions and therefore also not represented in the spectral line portion of the database for oxygen, however empirical absorption cross sections can be used to fill in the gap here.

#photo