RFM v4.3 Release Notes


After a longer-than-expected gestation period, RFM v4.3 is now ready to be released into the wild. It doesn't contain all the features I would have liked - these will have to be incorporated in updates - but the following is what it does have. In addition there are two new/modified F77 programs to convert HITRAN data to RFM-format input


You can use the new RFM v4.3 with existing driver tables, binary HITRAN files and cross-section files. There will be a slight difference in the results due to the new TIPS data interpolation, but this can be over-ridden with the new V42 Flag.

However, the numbering of some minor line molecules has been changed: molecules#39-46 have been reassigned (see RFM Absorbers) although it is unlikely that you required any of these for modelling the earth's atmosphere.

The numbering of all cross-section molecules has also been changed (RFM indices >50 are now all moved to >100 to allow for future expansion of the list of line molecules). But for the moment the RFM will accept either old or new designations so any RFM .xsc files with the old IDs will still be understood (eg an .xsc file containing molecule ID either 51 or 111 will be interpreted as CFC-11.

rfmsiz.inc This is the file containing most of the array size parameters used by the RFM and which you will probably have had to edit depending on your own requirements. It's probably easier to start again with the new version that comes with the RFM v4.3 delivery but if you really want to keep your old file and its parameter values you'll need to make the following changes

Also, there are a vast number of new HNO3 microwave lines (<50 cm-1) and COF2 lines in the 770cm-1 (13µm) region, which are likely to result in the (mid-run) fatal error message:
F-REACYC: Cyclic line buffer full (MAXCYC too small)
The default value MAXCYC = 10000 is adequate for other molecules (and previous versions of HITRAN) but if you are modelling microwave HNO3 a value of MAXCYC = 45000 is required, and for 13µm COF2 MAXCYC = 30000.

If you're trying to use the separate line parameter file for SF6 (30_hit08.par) - which has nearly 3 million transitions from 580-996 cm-1, i.e. half the size of the entire HITRAN 2012 database - you'll need MAXCYC=500 000.

If that doesn't fix the error (depending on which other molecules you are also modelling) then try increments of 5000 until it does.

Future Developments

The list of things still to be incorporated into future updates of the RFM