search

AER-RC / LNFL

Binary Line File Generator by Atmospheric and Environmental Research
Other
8 stars 5 forks source link

example TAPE5 files and/or utilities to generate TAPE5 files? #13

Open BenjaminTJohnson opened 1 year ago

BenjaminTJohnson commented 1 year ago

Hi,

I'm working on learning LNFL (finally) and would like to see some example TAPE5 files that work with LNFL. In particular, I'm interested in examples (with angular dependence) for UV, VIS, near-IR, IR, far-IR, and MW. Ultimately this will be used for training transmittance regression coefficients in CRTM.

The old TAPE5 files (~10 years old) that we have generate an error mesage forrtl: severe (64): input conversion error, unit 5, file /data/users/bjohnson/CRTM/LNFL/TAPE5 -- so something is wrong with the formatting.

Also, do TAPE5 files for LBLRTM work with LNFL (this might be the problem is that I'm confusing the two).

pernak18 commented 1 year ago

@BenjaminTJohnson have a look at https://github.com/AER-RC/LNFL/issues/11#issuecomment-1479687046 and let us know if that helps clarify.

no, LBLRTM and LNFL TAPE5 files are NOT interchangeable. LNFL TAPE5s are much simpler.

thanks for you interest in LNFL.

BenjaminTJohnson commented 1 year ago

@pernak18 Thanks Rick -- I'm looking for a robust example, are there some more complex example TAPE5 tables that I can download / look at? preferably with some explanation. For example, here's one I found, but I can't interpret all fields: 

TAPE5 FOR MLS 
0        1         2         3         4         5         6         7         8         9
123456789-123456789-123456789-123456789-123456789-123456789-123456789-123456789-123456789-
$ STANDARD MID-LATITUDE SUMMER ATMOSPHERE
HI=1 F4=1 CN=1 AE 0 EM=0 SC=0 FI=0 PL=0 TS=0 AM=0 MG=1 LA=0    1        00   00    0    0   14
294.2
        0
     2500.
  1 51    7  1.000000MIDLATITUDE SUMM H1=    0.00 H2=   70.00 ANG=   0.000 LEN= 0
    952.1147        291.77              3   0.000 1013.00 294.20  1.100  891.46 289.25
   1.5946558E-02  3.5495765E-04  3.1872162E-08  3.2014773E-07  1.4735235E-07  1.7007853E-06  2.0897518E-01  2.0212141E+24
    841.8970        287.03              3                         2.100 792.287 284.60
   1.1230157E-02  3.5495876E-04  3.5456235E-08  3.2014808E-07  1.4203219E-07  1.7007861E-06  2.0897572E-01  1.6594377E+24
    755.3917        282.23              3                         2.900 718.704 279.80
 ...

in particular I'm not sure what the "14" , "294.2" and "0 2500" are? Also, the values starting with " 3 ..." in the profile.

pernak18 commented 1 year ago

@pernak18 Thanks Rick -- I'm looking for a robust example, are there some more complex example TAPE5 tables that I can download / look at? preferably with some explanation. For example, here's one I found, but I can't interpret all fields:

TAPE5 FOR MLS 
0        1         2         3         4         5         6         7         8         9
123456789-123456789-123456789-123456789-123456789-123456789-123456789-123456789-123456789-
$ STANDARD MID-LATITUDE SUMMER ATMOSPHERE
HI=1 F4=1 CN=1 AE 0 EM=0 SC=0 FI=0 PL=0 TS=0 AM=0 MG=1 LA=0    1        00   00    0    0   14
294.2
        0
     2500.
  1 51    7  1.000000MIDLATITUDE SUMM H1=    0.00 H2=   70.00 ANG=   0.000 LEN= 0
    952.1147        291.77              3   0.000 1013.00 294.20  1.100  891.46 289.25
   1.5946558E-02  3.5495765E-04  3.1872162E-08  3.2014773E-07  1.4735235E-07  1.7007853E-06  2.0897518E-01  2.0212141E+24
    841.8970        287.03              3                         2.100 792.287 284.60
   1.1230157E-02  3.5495876E-04  3.5456235E-08  3.2014808E-07  1.4203219E-07  1.7007861E-06  2.0897572E-01  1.6594377E+24
    755.3917        282.23              3                         2.900 718.704 279.80
 ...

in particular I'm not sure what the "14" , "294.2" and "0 2500" are? Also, the values starting with " 3 ..." in the profile.

@BenjaminTJohnson that is definitely not an LNFL TAPE5. it's an LBL one

BenjaminTJohnson commented 1 year ago

Interesting. LNFL runs this without reporting any error, and produces TAPE3, TAPE6, and TAPE10 files. I'm happy to dig through more detailed documentation with examples (preferably).

pernak18 commented 1 year ago

@kcadyper do you know why LNFL would run with an LBL TAPE5?

@BenjaminTJohnson LNFL exists strictly to convert ASCII spectroscopy to a binary file readable to LBL, so all you need to specify is a spectral range and what molecules for which you want line parameters (there are other options, like broadening, speed dependence, and line coupling if the user wants them), not the layer information (P, T, z, and concentration) you provided because no RT is being calculated in LNFL. this is why i said the LNFL TAPE5 is much simpler

BenjaminTJohnson commented 1 year ago

Okay, sorry for the confusion. I'm just trying to get an end-to-end pathway for generating transmittance profiles for typical atmospheric constituents for fast atmospheric radiative transfer (CRTM specifically). If there are precomputed inputs for LBLRTM or MonoRTM, I can skip the LNFL step...

kcadyper commented 1 year ago

Hi, Ben. The LNFL release has several examples of LNFL TAPE5 files in the run_example subdirectory. We don't provide the LNFL output file (TAPE3) because it's a binary file (and thus not very portable) and very large. I encourage you to also look at the run_examples subdirectory in the LBLRTM release. In particular look the simples input files (TAPE5s) (e.g., user_defined and built_in) and compare with the descriptions of the inputs in the LBLRTM instructions, also provided with the release.

meteorids commented 2 weeks ago

Hi,

I created the following LNFL TAPE5 file (see below) and would like LNFL to create the TAPE7 file but it does not. When I run LNFL it creates the correct files except the TAPE7.

$ First test TAPE5
500.000 2500.000 11111111111111111111111111111111111111111111111LNOUT

Could you please provide some assistance what the issue might be? Best regards

pernak18 commented 2 weeks ago

@jrmascio @kcadyper any guidance on the post from @meteorids ? i tried their TAPE5 in /nas/project/rc/rc2/rpernak/temp and reproduced the problem of no TAPE7 being produced. now i just don't understand why that would be, given the contents of the TAPE5

kcadyper commented 1 week ago

LNFL reads the TAPE5 as a fixed format file. v1 should occupy the first ten spaces on the line, and v2 the next ten. The TAPE5 from meteorids has v1 extending into the v2 field, so LNFL reads V2 as zero. Make sure the frequencies are in the proper fields, and LNFL will run to completion.

meteorids commented 1 week ago

Thank you for your response. My actual TAPE5 file looks like the one below with the correct formatting. Not sure what went wrong when I copy pasted it the first time. When I run LNFL with this file it still does not create a TAPE7 file. Please advise. 

$ First test TAPE5                                                      
   500.000  2500.000
11111111111111111111111111111111111111111111111LNOUT

I appreciate your assistance!