Negative optical depth exists

[WinKevin66]

Hi,
I hope you are OK.
When I try to get gas absorption optical depth, I found that there have some negative values in some wavenumber and layers.
The TAPE5 file is attached, using O2-A band(12738.000-13422.000), select no any continuum (ICNTNM = 0), the LNFL just use O2 with LNOUT EXBRD, some result is just like figure 1: first is wavenumber and then are each layers' value.
And if I change ICNTNM = 6, and select only 'XO2CN O2 continuum absorption multiplicative factor', there also have negative values.
The zip includes TAPE5 file and my result.
So, what should I do?
Thank you for your reply!
WinKevin

zip.tar.gz

[mjiacono]

You can filter the negative optical depth values in your result by setting the input parameter DPTMIN on Record 1.3 in the TAPE5 file. You may either set DPTMIN to the minimum acceptable OD value (below which lines will be rejected), or you may set DPTMIN to a negative number and the minimum allowable OD value will default to 0.0002.