Perturbing ASCII line files in Line_Files_By_Molecule to build TAPE3

[gerbteam]

LNFL v3.2
AER line file v3.8.1 (downloaded from Zenodo 09/2025)

We are looking to build a TAPE3 output file from LNFL with the AER v3.8.1 for input into LBLRTM but using a perturbed water vapour line file. Line files by molecule type were provided with a slightly older LNFL download tar package (from 2020 I believe). So having perturbed one of these, and creating a soft link to "TAPE2", the LNFL is run.

From the documentation provided with LNFL, I think the approach would be to:

1. Disable H2O in the TAPE5 "RECORD 3" for M1 (i.e. M1=0)
2. Add MRG2 to "RECORD 3"
3. Set M1=1 in RECORD 4

So my "standard" (using the AER line file) LNFL TAPE5 file looks like:

** START OF TAPE5 **
$ f100 format
0. 2500.
11111111111111111111111111111111111111111111111 LNOUT EXBRD
%%%%%%%%%%%%%%%%%%
12345678901234567890123456789012345678901234567890123456789012345678901234567890

** END OF TAPE5 **

The new TAPE5 to hopefully use a perturbed H2O file:

** START OF TAPE5 **
$ f100 format
0. 2500.
01111111111111111111111111111111111111111111111 LNOUT EXBRD MRG2
10000000000000000000000000000000000000000000000
12345678901234567890123456789012345678901234567890123456789012345678901234567890

** END OF TAPE5 **

Is this the correct approach - appreciate you cannot comment on what is done to the H2O line file (line_files_By_Molecule_type/01_H2O/01_H2O), but assuming that is edited correctly, will everything else be ok?

Thanks!

[gerbteam]

Also, just to elaborate "will everything else be ok" - here I'm thinking about consistency with the broadening parameters and the water vapour continuum.

[matt-alvarado]

What you have should work, but to be extra safe, make the number of digits in the "1000..." line the same as the number of digits in the "0111..." line. Let us know if you see any problems or error messages

[gerbteam]

What you have should work, but to be extra safe, make the number of digits in the "1000..." line the same as the number of digits in the "0111..." line. Let us know if you see any problems or error messages

Many thanks - yes, the number of digits are exactly the same already. A couple of test runs with LBLRTM are being done now by the student - but I'll report back if any issues.

Thanks - please leave issue open for the moment if that's ok.