Actinic flux and airmass factor in Cloud-J module

[mcdonh1718]

Your name

Helena McDonald

Your affiliation

MIT

Please provide a clear and concise description of your question or discussion topic.

I'm working on a change to the chemical mechanism so that we can have full chemistry in the mesosphere (above 50km) rather than the simplified version that geoschem currently uses. My implementation is currently having an issue with an excess of ozone at the top layer of the model:

My working theory is that the photolysis module isn’t properly accounting for the solar beam attenuation through the atmosphere above 80km (because it's not directly simulated), meaning the incident actinic flux is too high and ozone is being over-produced. I’m trying to find a means of reducing the actinic flux by accounting for the increased optical depth and airmass factor introduced by the not-simulated air above the gridbox, and I’m hoping to leverage existing functions in CloudJ as opposed to hard-coding anything.

I’m hoping I can get your help interpreting some of the functions in CloudJ to see if they do what I actually need.

The biggest problem seems to be the airmass factor; for solar zenith angles below 75 I’m fine using 1/U0, but between 75 and 98 I need to use solutions to the chapman function (air mass factor!) which account for the sphericity of the atmosphere and refraction. I think the SPHERE1R function probably does this, but I’m a little confused with the definitions:

The way the AMF matrix is defined is confusing; if I pull AMF(72, 73), do I get the airmass factor between layer 72 and the top of the atmosphere? I'm thinking the solution here would be to manually adjust the

There’s also some discussion later in the OPMIE code that seems to act as though AMF can be negative, which is quite confusing:

This documentation also makes me curious about how the above-CTM top-of atmosphere height is set. Based on this explanation, it seems that’s determined by ZHL(L1U+1) where a ZZHT number of centimeters is stacked on top of the top-of-CTM radius. ZZHT is set as 5.0d5 = 50000cm = 5km. I’m confused by how ZZHT is labeled as a scale height but used as an absolute measure in cm in this application, and I'm also curious why 5km was picked as the value. I was thinking that a very straightforward attempt to resolve this could be to just increase the value of zzht. Using the standard equation for atmospheric scale height and the air density there, the actual scale height would be closer to 15km.

Optical depth is another confusing one; I can’t tell if there is a calculation for optical depth that includes the air above the gridbox, or if it’s just layer-to-layer. DTAUX seems to just calculate the optical depth of each layer, not anything above it. If that’s the case I don’t understand how that differs from DTAU, which is the local optical depth for each CTM level. Is there a subtlety here I’m missing? Then there’s TTAU and FTAU which are the vertical OD and attenuation respectively; I think possibly what I’d be looking for here is FTAU, where the solar beam has been attenuated passing through to the layer I need it at?

I know this is a lot of questions - I'm sort of throwing things at the wall to see what sticks right now!

[yantosca]

Tagging @lizziel who is our Cloud-J developer.

[lizziel]

Hi @mcdonh1718, I will transfer this issue to the Cloud-J repository issues page. I am also going to bring in Michael Prather who is the developer of Cloud-J. @pratherUCI.

[lizziel]

@pratherUCI - Michael, would you be able to take a look at Helena's questions about the Cloud-J implementation? I suspect you will have ready answers.

[pratherUCI]

Dear Helena,

I will give you some answers and a reading lesson. Then you can rethink the next set of questions.

#1 reading: you really need to read carefully to see the difference between spherical and refracting atmospheres and how AMF is used: Prather, M.J. and J.C. Hsu (2019) A round Earth for climate models, Proc Natl Acad Sci, 116 (39) 19330-19335; https://doi.org/10.1073/pnas.1908198116

#2 Each layer is defined by the column mass of O2, O3, N2, etc in molec per cm2, which is the vertical column. The airmass factor (AMF) is the solar ray's slant path distance relative to the vertical (overhead) distance. The sum of the AMFs x colm density over all layers is the # molecules per cm2 in the path back to the sun, and that times the cross section (weighted by temperature and pressure along that path) give the optical depth.

#3 Spherical atmospheres will be important for mesosphere, but Refraction may be overkill since (read about it) refraction only has a notable difference if the ray path comes close to the surface.

#4 For a flat atmosphere the dZ of each layer is not important, but for spherical and above, the geometric height of the layer is important for the AMF. The ZZHT is the geometric thickness of the uppermost top layer since it cannot be reliably calc from hydrostatic equation.because the uppermost layer is the mass above the CTM top, which of course has no T or other data, including O3, (but it has O2 as 21% of the mass!). The ZZHT value will only matter for the spherical. SO run you code as a plat atmosphere to test you O3 problem. I doubt it has anything to do with Cloud-J.. Do not forget, you should not do chemistry or photochemistry on this topmost layer.

#5 From the Cloud-J notes: the J-values are not very good above 64 km, that is inherently part of the design and optimization of Cloud-J.

Keep going, Michael

[mcdonh1718]

Thanks for the detailed response! One clarifying question for point #4: why is it important not to do chemistry/photochemistry on the topmost layer? This might be the crux of my issue; I've been trying to get chemistry to work over the full model grid.