RFM File Formats

.tab File: Tabulated Absorption Coefficient File

05NOV17

This describes the format used with RFM v4.33 onwards. Older format is described [HERE]

Type
Output(s), if TAB Flag enabled.
Optional Input(s) if LUT Flag enabled (alternatively use .svd files).

Description
Values of ln(k), where k is the Absorption Coefficients [m2/kmol] (1 kmol approx 6E26 molec) tabulated as function of wavenumber, pressure, temperature and (optionally) VMR. A value of -99.0 is used for ln(k) if k indistinguishable from zero at single precision.

Filename(s)
Default is tab_[s][g].asc where [s] is the spectral range and [g] is the gas
Different names may be specified in the *TAB section of the Driver Table.

Structure
!HEADER1
!HEADER2
!HEADER3
RFMT
ID[.ISO] NV V1 V2 DV NA NP NT NQ
P(1) P(2) ... P(NP)
T_p(1) T_p(2) ... T_p(NP)
V_p(1) V_p(2) ... V_p(NP)
T(1) T(2) ... T(NT)
Q(1) Q(2) ... Q(NQ)
V(1) K(1,1,1,1), K(2,1,1,1)... K(NQ,NT,NP,1)
V(2) K(1,1,1,2), K(2,1,1,1)... K(NQ,NT,NP,2)
...
V(NV) K(1,1,1,NV), K(2,1,1,NV)... K(NQ,NT,NP,NV)
[eof]

Fields
FieldTypeDescriptionActual Values
!HEADER1 C*80 Spectrum type, ray-path and RFM version ID
!HEADER2 C*80 Text read from *HDR section of Driver Table
!HEADER3 C*80 Captions for fields in record after next
RFMT R*4 File format identifier 1.0
ID[.ISO] A5 Absorber/isotopologue identifier consisting of...
ID I HITRAN Index of absorber See List of Absorbers
.ISO I (optional) '.' then HITRAN Isotope# See List of Isotopes
NV I No. of Spectral Points >0 if [cm-1], <0 if [GHz]
V1 R*8 First/Lowest Spectral pt [cm-1]/[GHz] ≥0
V2 R*8 Last/Highest Spectral pt [cm-1]/[GHz] ≥V1
DV R*8 Nominal/min spectral interval [cm-1]/[GHz] ≥0
NA I Tot no K points at each spectral pt NP * NT * NQ
NP I No. of Points on Pressure axis >0
NT I No. of Temperature Points >0 (Abs) or <0 (Rel)
NQ I No. of VMR Points >0 (Abs) or <0 (Rel)
P R*4 Pressure axis values [mb] >0
T_p R*4 Temperature profile [K] >0
V_p R*4 VMR profile [ppmv] ≥0
T R*4 Temperature axis values [K] >0 [Abs], ± [Rel]
Q R*4 VMR axis values [%] >0
V R*8 Spectral axis value [cm-1]/[GHz] V1 ≤ V ≤ V2
K R*4 K(NQ,NT,NP,NV) ln(Absorption Coeff) [m2/mole] -99.0 for k=0

Notes
  1. This new format is designed to be more flexible, assuming arbitrarily spaced axes in all domains, and adding the option of partial pressure dependence (where self-broadening is likely to be important). In fact the temperature dependence is also optional.
  2. The first 3 HEADER records .tab files are similar to RFM Spectral Files
  3. Records 1-5 are all written as 80-character records, which simplifies reading binary files
  4. The 4th record contains a single real number RFMT, currently 1.0, to identify the format. This is to build in some robustness to any future changes of the file format, with the intention that older formats remain supported
  5. The 5th record gives the molecule ID, wavenumber range and array sizes.
    • If no isotopologue distinction is applied, the A*5 field just contains the RFM molecule index. If particular isotopologues are specified it contains a real number with molecule index as the integer part and Isotopologue ID as the decimal part.
    • If NV is negative this indicates that the spectral limits V1,V2 and the file contains |NV| spectral points V specified in [GHz] rather than [cm-1]
    • If NT is negative this indicates that a relative temperature axis of |NT| points is being used
    • DV, for regularly spaced files (e.g. as generated by the RFM) is the grid spacing. For irregularly spaced files it is the mininum grid spacing, or the original grid spacing before the file was spectrally compressed (the actual value is not used by the RFM when the file is read in).
  6. Then follows one or more records containing NP pressure axis values P [mb or hPa] in monotonically increasing or decreasing order
  7. Then follows one or more records containing NP Temperature profile values T_p corresponding to each P. This is only useful if a relative temperature axis is being used, but always included to keep the file format consistent.
  8. Then follows one or more records containing NP VMR profile values V_p corresponding to each P. This is only useful if the absorption coefficient is also a function of VMR, i.e., NQ >1, but again always included for consistency.
  9. Then follows one or more records containing |NT| temperature axis values [K], either absolute (NT>0) or relative to T_p (NT<0) in monotonically increasing or decreasing order.
    • If NT=1 then the K values are all computed for a single temperature value T(1).
    • If NT=-1 then the K values are computed for a single pressure-dependent temperature value T_p + T(1) (you would probably set T(1) = 0 in this case).
    • The advantage of using a relative temperature grid, eg T = -30, -20 ... +30, is that you can capture the range of atmospheric variability with a smaller look-up table (NT=7) as opposed to using absolute temperatures say T = 180, 190, ... 310 (NT=14) with large parts of the table never used (eg 300K at the tropopause)
  10. Then follows one or more records containing NV volume mixing ratio axis values [%], which are used to scale the VMR profile V_p
    • If NQ=1 then Q(1) = 100 is the expected value, meaning that K values for each P value were computed using the VMR profile values V_p
  11. When creating isotopic .tab files (see Notes on use of the TAB Flag) a separate file is created for each isotopic profile supplied in the *ATM section. These have "iN" where N is the isotope number, inserted into the output filename (eg a profile *CH4(2) will result in a .tab filename ...ch4i2...). There will also be a file "i0" for remaining, unspecified isotopes (which may contain all zeros if all isotopes were specified - consider this a "feature" of the RFM). This information is also inserted into the [.ISO] as a decimal value after the IDX value, eg "06.2" and "06.0".
  12. When using .tab files as input to RFM calculations, if no isotopic profile is specified the RFM will only accept a generic .tab file for that molecule (eg with the ID " 6" rather than " 6.x"). Conversely, if isotopic profiles are specified the RFM will only accept .tab files for that particular isotope (eg with the ID "6.2") and any unspecified isotopes will be modelled using the default isotopic file (eg "6.0"). , Note: if, for example, you generate .tab files for the first two isotopes (6.1, 6.2) the remaining isotopes will be included in the file with ID 6.0. However, if you then use these as input to the RFM specifying only the first isotope, 6.1, as a separate profile, the RFM will incorrectly assume that isotope#2 is included in the default file 6.0 and ignore the file 6.2.
  13. When using .tab files based on irregular spectral grids as input, the grid information is automatically used by the RFM to linearly interpolate the data to the RFM internal grid (which may be regularly spaced). This is distinct from using the GRD Flag which specified an irregularly spaced RFM internal grid.

Example
 ! o3      Tabulated Absorp.Coeff. created by RFM v.3.42_OXF
 !  TEST
 !NL  NV    V1      DV    NP   P1         DP         NT       T1          DT
O3__0053  3 LIN
 0  2701 1036.0250 0.0005 20  -6.907755   0.727132    8     120.000      30.000
   22.20402       26.26059       26.64072       19.32595       11.47375
   6.571761       3.676178       1.906070      0.9411216      0.4572936
...
[eof]