Re: [arts-users] Use of scattering methods
Hi Patrick, Thank you for your help. It is indeed very difficult to achieve inversion of temperature and humidity profiles under cloud conditions. I have another question, is it in the planetary toolbox that the best way to use it is iyFOS or iyDOTIT? IyMC seems too slow. At 2023-11-06 15:52:08, "Patrick Eriksson" wrote: >Hi, > >Very hard to answer without more specific information what you want to do. > >If you want to do OEM-type inversion involving scattering, then iyHybrid >is the only option inside ARTS. It is reasonable fast, if used with e.g. >DISORT. > >DISORT is the fastest scattering solver inside ARTS, but has several >limitations. > >iyMC is the slowest, but still the one to use for 3D calculations. > >To retrieve atmospheric temperature or humidity with scattering present >is a very hard problem. > >Bye, > >Patruick > > > >On 2023-11-02 13:44, suifengbenpao2023 wrote: >> Hi Patrick, >> >> Thanks again for your help! Is the iymc method generally slower than the >> iydoit method? If the radiative transfer rate calculated by the scattering >> method is very slow, does it mean that it is difficult to invert the >> atmospheric temperature and humidity profile from the scattering scene? Is >> there a quick way to calculate scattering? >> >> Looking forward to your reply again, thanks! >> >> Sincerely， >> >> Jiaan He. >> >> At 2023-11-02 16:15:43, "Patrick Eriksson" >> wrote: >>>Hi, >>> >>>In general, it works perfectly fine to use PlanckBT from start. But >>>could depend on the calibration of the instrument you simulate. >>> >>>There is one exception, and that's iyMC. It does not accept PlanckBT, >>>just RJBT. >>> >>>Bye, >>> >>>Patrick >>> >>> >>>On 2023-11-02 08:42, suifengbenpao2023 wrote: Hi Patrick, Thank you for your help! Based on your suggestion, I will no longer use iyfos and instead use iyMC to handle scattering situations. But I have another question that confuses me. I often see routines that use StringSet (iy_unit, "1") for unit conversion before calculating scattering, and then ultimately convert to StringSet (iy_unit, "PlanckBT"). Why do we do this? Can't we just switch to Planck brightness temperature? Looking forward to your reply again, thanks! Sincerely， Jiaan He. At 2023-11-02 14:17:07, "Patrick Eriksson" wrote: >Hi, > >FOS is "dead" and no plans to get it back to live. > >A main reason for starting FOS was to have scattering calculations, >providing the Jacobian. This was newer completed. Instead, also iyHybrid >was started and that method provides a Jacobian (not the totally full >one, but seems to be good enough). > >Bye, > >Patrick > >On 2023-11-02 01:20, suifengbenpao2023 wrote: >> Dear ARTS community, >> >> I am dealing with scattering situations and have used the iyFOS method >> in Arts-2.2 version, but there is no FOS method in Arts-2.4. May I ask >> if the method name has changed? I know and have consulted relevant >> literature on MC and DOIT methods, but is there any reference to FOS >> methods? >> >> Looking forward to your reply！Thank you! >> >> Best wishes, >> >> Jiaan He
Re: [arts-users] Use of scattering methods
Hi, Very hard to answer without more specific information what you want to do. If you want to do OEM-type inversion involving scattering, then iyHybrid is the only option inside ARTS. It is reasonable fast, if used with e.g. DISORT. DISORT is the fastest scattering solver inside ARTS, but has several limitations. iyMC is the slowest, but still the one to use for 3D calculations. To retrieve atmospheric temperature or humidity with scattering present is a very hard problem. Bye, Patruick On 2023-11-02 13:44, suifengbenpao2023 wrote: Hi Patrick, Thanks again for your help! Is the iymc method generally slower than the iydoit method? If the radiative transfer rate calculated by the scattering method is very slow, does it mean that it is difficult to invert the atmospheric temperature and humidity profile from the scattering scene? Is there a quick way to calculate scattering? Looking forward to your reply again, thanks! Sincerely， Jiaan He. At 2023-11-02 16:15:43, "Patrick Eriksson" wrote: Hi, In general, it works perfectly fine to use PlanckBT from start. But could depend on the calibration of the instrument you simulate. There is one exception, and that's iyMC. It does not accept PlanckBT, just RJBT. Bye, Patrick On 2023-11-02 08:42, suifengbenpao2023 wrote: Hi Patrick, Thank you for your help! Based on your suggestion, I will no longer use iyfos and instead use iyMC to handle scattering situations. But I have another question that confuses me. I often see routines that use StringSet (iy_unit, "1") for unit conversion before calculating scattering, and then ultimately convert to StringSet (iy_unit, "PlanckBT"). Why do we do this? Can't we just switch to Planck brightness temperature? Looking forward to your reply again, thanks! Sincerely， Jiaan He. At 2023-11-02 14:17:07, "Patrick Eriksson" wrote: Hi, FOS is "dead" and no plans to get it back to live. A main reason for starting FOS was to have scattering calculations, providing the Jacobian. This was newer completed. Instead, also iyHybrid was started and that method provides a Jacobian (not the totally full one, but seems to be good enough). Bye, Patrick On 2023-11-02 01:20, suifengbenpao2023 wrote: Dear ARTS community, I am dealing with scattering situations and have used the iyFOS method in Arts-2.2 version, but there is no FOS method in Arts-2.4. May I ask if the method name has changed? I know and have consulted relevant literature on MC and DOIT methods, but is there any reference to FOS methods? Looking forward to your reply！Thank you! Best wishes, Jiaan He
Re: [arts-users] Use of scattering methods
Hi Patrick, Thanks again for your help! Is the iymc method generally slower than the iydoit method? If the radiative transfer rate calculated by the scattering method is very slow, does it mean that it is difficult to invert the atmospheric temperature and humidity profile from the scattering scene? Is there a quick way to calculate scattering? Looking forward to your reply again, thanks! Sincerely， Jiaan He. At 2023-11-02 16:15:43, "Patrick Eriksson" wrote: >Hi, > >In general, it works perfectly fine to use PlanckBT from start. But >could depend on the calibration of the instrument you simulate. > >There is one exception, and that's iyMC. It does not accept PlanckBT, >just RJBT. > >Bye, > >Patrick > > >On 2023-11-02 08:42, suifengbenpao2023 wrote: >> Hi Patrick, >> >> Thank you for your help! Based on your suggestion, I will no longer use >> iyfos and instead use iyMC to handle scattering situations. But I have >> another question that confuses me. I often see routines that use >> StringSet (iy_unit, "1") for unit conversion before calculating >> scattering, and then ultimately convert to StringSet (iy_unit, >> "PlanckBT"). Why do we do this? Can't we just switch to Planck >> brightness temperature? >> >> Looking forward to your reply again, thanks! >> >> Sincerely， >> Jiaan He. >> >> >> At 2023-11-02 14:17:07, "Patrick Eriksson" >> wrote: >>>Hi, >>> >>>FOS is "dead" and no plans to get it back to live. >>> >>>A main reason for starting FOS was to have scattering calculations, >>>providing the Jacobian. This was newer completed. Instead, also iyHybrid >>>was started and that method provides a Jacobian (not the totally full >>>one, but seems to be good enough). >>> >>>Bye, >>> >>>Patrick >>> >>>On 2023-11-02 01:20, suifengbenpao2023 wrote: Dear ARTS community, I am dealing with scattering situations and have used the iyFOS method in Arts-2.2 version, but there is no FOS method in Arts-2.4. May I ask if the method name has changed? I know and have consulted relevant literature on MC and DOIT methods, but is there any reference to FOS methods? Looking forward to your reply！Thank you! Best wishes, Jiaan He
Re: [arts-users] Use of scattering methods
Hi, In general, it works perfectly fine to use PlanckBT from start. But could depend on the calibration of the instrument you simulate. There is one exception, and that's iyMC. It does not accept PlanckBT, just RJBT. Bye, Patrick On 2023-11-02 08:42, suifengbenpao2023 wrote: Hi Patrick, Thank you for your help! Based on your suggestion, I will no longer use iyfos and instead use iyMC to handle scattering situations. But I have another question that confuses me. I often see routines that use StringSet (iy_unit, "1") for unit conversion before calculating scattering, and then ultimately convert to StringSet (iy_unit, "PlanckBT"). Why do we do this? Can't we just switch to Planck brightness temperature? Looking forward to your reply again, thanks! Sincerely， Jiaan He. At 2023-11-02 14:17:07, "Patrick Eriksson" wrote: Hi, FOS is "dead" and no plans to get it back to live. A main reason for starting FOS was to have scattering calculations, providing the Jacobian. This was newer completed. Instead, also iyHybrid was started and that method provides a Jacobian (not the totally full one, but seems to be good enough). Bye, Patrick On 2023-11-02 01:20, suifengbenpao2023 wrote: Dear ARTS community, I am dealing with scattering situations and have used the iyFOS method in Arts-2.2 version, but there is no FOS method in Arts-2.4. May I ask if the method name has changed? I know and have consulted relevant literature on MC and DOIT methods, but is there any reference to FOS methods? Looking forward to your reply！Thank you! Best wishes, Jiaan He
Re: [arts-users] Use of scattering methods
Hi Patrick,Thank you for your help! Based on your suggestion, I will no longer use iyfos and instead use iyMC to handle scattering situations. But I have another question that confuses me. I often see routines that use StringSet (iy_unit, "1") for unit conversion before calculating scattering, and then ultimately convert to StringSet (iy_unit, "PlanckBT"). Why do we do this? Can't we just switch to Planck brightness temperature? Looking forward to your reply again, thanks! Sincerely， Jiaan He. At 2023-11-02 14:17:07, "Patrick Eriksson" wrote: >Hi, > >FOS is "dead" and no plans to get it back to live. > >A main reason for starting FOS was to have scattering calculations, >providing the Jacobian. This was newer completed. Instead, also iyHybrid >was started and that method provides a Jacobian (not the totally full >one, but seems to be good enough). > >Bye, > >Patrick > >On 2023-11-02 01:20, suifengbenpao2023 wrote: >> Dear ARTS community, >> >> I am dealing with scattering situations and have used the iyFOS method >> in Arts-2.2 version, but there is no FOS method in Arts-2.4. May I ask >> if the method name has changed? I know and have consulted relevant >> literature on MC and DOIT methods, but is there any reference to FOS >> methods? >> >> Looking forward to your reply！Thank you! >> >> Best wishes, >> >> Jiaan He
Re: [arts-users] Use of scattering methods
Hi, FOS is "dead" and no plans to get it back to live. A main reason for starting FOS was to have scattering calculations, providing the Jacobian. This was newer completed. Instead, also iyHybrid was started and that method provides a Jacobian (not the totally full one, but seems to be good enough). Bye, Patrick On 2023-11-02 01:20, suifengbenpao2023 wrote: Dear ARTS community, I am dealing with scattering situations and have used the iyFOS method in Arts-2.2 version, but there is no FOS method in Arts-2.4. May I ask if the method name has changed? I know and have consulted relevant literature on MC and DOIT methods, but is there any reference to FOS methods? Looking forward to your reply！Thank you! Best wishes, Jiaan He
[arts-users] Use of scattering methods
Dear ARTS community, I am dealing with scattering situations and have used the iyFOS method in Arts-2.2 version, but there is no FOS method in Arts-2.4. May I ask if the method name has changed? I know and have consulted relevant literature on MC and DOIT methods, but is there any reference to FOS methods? Looking forward to your reply！Thank you! Best wishes, Jiaan He
