Dear Stefan, Thank you so much for quickly getting back to me!
Looking through the example code and setting the line width cutoff has helped. Thank you so much for that! Hope you have a great day! Sincerely, Michie On 6/17/25, 10:22 AM, "Stefan Buehler" <stefan.bueh...@uni-hamburg.de> wrote: Dear Michie, do you want the absorption to be exactly 0 in the windows? That is unphysical, but you can achieve it in ARTS by setting a line width cutoff. I guess that is what HAPI does as well. You can use the workspace methods with „Cutoff“ in their name for this (there are a few variants, depending on whether you want to set different cutoffs for different species, or use the same value for all). The recent ARTS paper has some discussion on line shape functions in general. It also has a code example setting a cutoff. Cutoffs are mainly used for computational efficiency and (in the case of water vapor) to have a well defined separation between line spectrum and absorption continuum. Best wishes, Stefan On 17 Jun 2025, at 16:34, De Vera, Michie Vianca Dichaves via arts_users.mi wrote: > Hi Oliver, > > Many thanks for your quick response! > > Here is the link to Bohren and Clothiaux (2006) : > https://doi.org/10.1002/9783527618620.ch2 > > Figure 2.12 on page 93 shows the absorption cross section for ozone. > > I’m also attaching a plot I made using the HITRAN python package HAPI. I > guess since HITRAN doesn’t have lines available for some regions, HAPI just > sets the values to 0 in those regions, but ARTS seems to fill those values > in, as you’ve shown in the plot you sent. Is there a way to go about this or > will I have to postprocess the results I’m getting? > > Thank you so much! > > Sincerely, > Michie > > From: Oliver Lemke <ole...@gmail.com<mailto:ole...@gmail.com>> > Date: Tuesday, June 17, 2025 at 6:44 AM > To: De Vera, Michie Vianca Dichaves > <mdeve...@illinois.edu<mailto:mdeve...@illinois.edu>> > Cc: ARTS Users List > <arts_users.mi@lists.uni-hamburg.de<mailto:arts_users.mi@lists.uni-hamburg.de>> > Subject: Re: [arts-users] Problem with ozone absorption coefficients > Hi Michie, > > It looks like there are just no lines available from HITRAN (which our > catalog is based on) in that region (I attached a zoomed in plot in case > others want to chime in). Can you link to a plot that shows the feature > you're expecting? > > Cheers, > Oliver > >> On 16. Jun 2025, at 19:30, De Vera, Michie Vianca Dichaves via arts_users.mi >> <arts_users.mi@lists.uni-hamburg.de<mailto:arts_users.mi@lists.uni-hamburg.de>> >> wrote: >> >> Good day! >> I am new to ARTS and have been using the newest version on Python. I’m >> trying to plot the absorption cross section for ozone as a function of >> wavelength. However, I noticed that I’m only getting some of the features >> right. As seen in the image attached, for example, there should be a dip in >> the values around 6.3 microns based on textbook images, but I don’t see that >> here. Other wavelengths should also have dips that are not seen in the >> image. I have tried to increase the spectral resolution, but I still do not >> get these features. >> Below is part of the Python code I’m using to get the absorption >> coefficient values. I tried to follow one of the example files, but I would >> appreciate any help to see if I’m missing something. >> ws = pyarts.workspace.Workspace() >> ws.abs_speciesSet(species=[‘O3’]) >> ws.abs_lines_per_speciesReadSpeciesSplitCatalog(basename="/dir/arts/arts-cat-data-2.6.14/lines/") >> ws.lbl_checkedCalc() >> ws.propmat_clearsky_agendaAuto() >> ws.jacobian_quantities = [] >> ws.select_abs_species = [] >> ws.f_grid = frq_grid >> ws.rtp_mag = [] >> ws.rtp_los = [] >> ws.rtp_pressure = 100681 >> ws.rtp_temperature = 285.77 >> ws.rtp_nlte = pyarts.arts.EnergyLevelMap() >> ws.rtp_vmr = [0.2676793e-07] >> ws.stokes_dim = 1 >> ws.AgendaExecute(a=ws.propmat_clearsky_agenda) >> Thank you so much! >> Sincerely, >> Michie<arts_O3.png> > [cid:484ab79a-4622-4a09-a516-7c1085fae14c@namprd11.prod.outlook.com<mailto:484ab79a-4622-4a09-a516-7c1085fae...@namprd11.prod.outlook.com>]
