Definitions of variables

[tim]

Hello!

I am processing climate data produced by CESM1.2 / CAM4. My goal is to provide visualizations of the top-of-atmosphere radiation budget to assess the impact of localized surface albedo modification. I am an intern who is new to CESM, and my question is very simple.

Where are the model output variables rigorously defined?

Fields such as FLNT : Net longwave flux at top of model are partially described within the NetCDF metadata, but there are numerous conventions that are confusing to newcomers such as myself. For instance, the "net" descriptor apparently is positive-up when referring to longwave, yet positive-down when referring to shortwave radiation. I have also seen that the phrase "top of model" refers to a different pressure level than "top of atmosphere", which would be difficult to infer from the variable metadata alone.

I want to make sure I'm not making silly mistakes by using the wrong variables or sign conventions or missing important details. Yet I have not found a centralized set of rigorous variable definitions (i.e., unambiguously defining "longwave", "net", and "top of model") in the CESM User Guide, CAM4 Scientific Description, or even the Master Field Lists. Does a suitable glossary exist, or are these potentially confusing conventions and definitions assumed to be common knowledge?

If I am asking in the wrong forum or have missed a relevant post, please let me know. I am happy to provide more information about my specific use case, but it may not be relevant to this general question.

Thank you,
Tim

 

[brianpm]

There is no comprehensive list. Some of the conventions can be understood by looking at some of the additional metadata in the output files, but to understand that metadata, you will need to refer to the CF conventions documentation (CF Conventions Home Page).

Regarding the radiative fluxes, you are right that conventions are not obvious for newcomers. There is a distinction between "top of model" and "top of atmosphere"; the "top of atmosphere" fluxes are calculated as an additional layer above the top of the model are are included to provide a more appropriate comparison with satellite observations. The model's global energy budget should be made from the "top of model" fluxes, while comparisons with observations should use "top of atmosphere."

The net radiative flux at "top of atmosphere" is sometimes called RESTOA ("RES" for residual), and at "top of model" it is called RESTOM. These are typically diagnosed from model output as

RESTOM = FSNT - FLNT
RESTOA = FSNTOA - FLUT

The RHS of these are the net shortwave and longwave fluxes. They are computed as differences:

FSNT = SOLIN - FSUT
FSNTOA = SOLIN - FSUTOA
FLNT = FLUT - [FLDT]

FLDT is not usually output from the model, but FLNT and FLUT are. SOLIN is the insolation.

Clear-sky fluxes are derived in the same way.

For further details, you can look directly at the radiation code in cesm2_1_1/components/cam/src/physics/rrtmg/radlw.F90 and in radsw.F90. You might also look at the AMWG diagnostics code (AMWG Diagnostics Package | Where to Find the Code).

 

[tim]

Thank you, Brian! I appreciate those references.

 

[Jen]

There is only FLNT, not FLNTOA. I want the latter, how can I make it? Thank you.

 

[Jen]

There is no comprehensive list. Some of the conventions can be understood by looking at some of the additional metadata in the output files, but to understand that metadata, you will need to refer to the CF conventions documentation (CF Conventions Home Page).

Regarding the radiative fluxes, you are right that conventions are not obvious for newcomers. There is a distinction between "top of model" and "top of atmosphere"; the "top of atmosphere" fluxes are calculated as an additional layer above the top of the model are are included to provide a more appropriate comparison with satellite observations. The model's global energy budget should be made from the "top of model" fluxes, while comparisons with observations should use "top of atmosphere."

The net radiative flux at "top of atmosphere" is sometimes called RESTOA ("RES" for residual), and at "top of model" it is called RESTOM. These are typically diagnosed from model output as

RESTOM = FSNT - FLNT
RESTOA = FSNTOA - FLUT

The RHS of these are the net shortwave and longwave fluxes. They are computed as differences:

FSNT = SOLIN - FSUT
FSNTOA = SOLIN - FSUTOA
FLNT = FLUT - [FLDT]

FLDT is not usually output from the model, but FLNT and FLUT are. SOLIN is the insolation.

Clear-sky fluxes are derived in the same way.

For further details, you can look directly at the radiation code in cesm2_1_1/components/cam/src/physics/rrtmg/radlw.F90 and in radsw.F90. You might also look at the AMWG diagnostics code (AMWG Diagnostics Package | Where to Find the Code).

 

[Jen]

RESTOA = FSNTOA - FLUT, why isn't "RESTOA" equal to "FSNTOA - FLNT"?

 

[brianpm]

The "TOA" concept is somewhat flawed. In terms of energy conservation, the variables that matter are the "top-of-model" fluxes. The TOA variables, at least AFAIK, are for comparison with ERBE and possibly CERES fluxes. The RESTOA variable can be thought of as:

RESTOA = FSNTOA - FLUT

Because:
1. FSNTOA = SOLIN - FSUTOA
2. SOLIN is really the total irradiance including whatever downward long wave is included.
3. CESM does not write a FLUTOA, so we must assume FLUT ~ FLUTOA

This is my understanding of the fluxes. If anyone has a useful elaboration / correction, I am all ears.

If you want to make your own FLUTOA output, I think you can follow the same steps as are used for the other fluxes. The flux is calculated, it just does not have an output variable defined.

 

[Jen]

Thanks for your kind reply. I understand.This is very helpful to me.