I am trying to identify galxies hosting ANG. What is the best way to do it?
I was checking the data specifications but couldn't find any label (like "subhalo hosts AGN") in the groups catalogue.
Does it make sense to use the "GFM_AGNRadiation " column (from the PartType0 snapshot), and check wether it has a positive value?
Thank you!
Diego
Dylan Nelson
23 Sep '19
Hi Diego,
No the GFM_AGNRadiation is a bit tangential, it would be better to use the accretion rate onto the black hole itself.
You're right this isn't stored "per subhalo", so for a given subhalo you have to actually load its black hole particle. This isn't such a trouble though, since there are so few black hole particles in the entire simulation.
Once you've loaded the BHs, you can use the BH_Mdot to convert to a bolometric luminosity using a simple model like Eqn. 5 of Nelson+ (2019).
Note this is very simple, it neglects things like obscuration effects, short duty cycles, specific observational details like energy bandpasses. If you are comparing to observations, you may want to think carefully about these and what might need to be done to produce an "observationally realistic AGN criterion" from the accretion rate (or accretion rate history).
Diego Sotillo Ramos
23 Sep '19
Thank you, Dylan,
I think your suggestion will be enough for my purposes (I just need a binary output). I will try it and come back to you if I need more feedback ;)
Rui Zhou
3 Nov
Hi Dylan,
I'm also trying to identify AGN host galaxies. I'd like to know what can I do if I already calculated the bolometric luminosity? I read your paper bud didn't find such a criteria to tell if a black hole is an AGN based on this luminosity.
Rui
Dylan Nelson
3 Nov
I am not sure there is an agreed upon definition. I think most observational works would define an AGN as any detectable SMBH, i.e. any L_bol above some number. The exact number would depend on the particular survey/paper/etc. I'd suggest to just adopt a similar number as whatever observational paper you find closest to the topic you are studying. This should give a starting point.
Hi, Dylan,
I am trying to identify galxies hosting ANG. What is the best way to do it?
I was checking the data specifications but couldn't find any label (like "subhalo hosts AGN") in the groups catalogue.
Does it make sense to use the "GFM_AGNRadiation " column (from the PartType0 snapshot), and check wether it has a positive value?
Thank you!
Diego
Hi Diego,
No the
GFM_AGNRadiationis a bit tangential, it would be better to use the accretion rate onto the black hole itself.You're right this isn't stored "per subhalo", so for a given subhalo you have to actually load its black hole particle. This isn't such a trouble though, since there are so few black hole particles in the entire simulation.
Once you've loaded the BHs, you can use the
BH_Mdotto convert to a bolometric luminosity using a simple model like Eqn. 5 of Nelson+ (2019).Note this is very simple, it neglects things like obscuration effects, short duty cycles, specific observational details like energy bandpasses. If you are comparing to observations, you may want to think carefully about these and what might need to be done to produce an "observationally realistic AGN criterion" from the accretion rate (or accretion rate history).
Thank you, Dylan,
I think your suggestion will be enough for my purposes (I just need a binary output). I will try it and come back to you if I need more feedback ;)
Hi Dylan,
I'm also trying to identify AGN host galaxies. I'd like to know what can I do if I already calculated the bolometric luminosity? I read your paper bud didn't find such a criteria to tell if a black hole is an AGN based on this luminosity.
Rui
I am not sure there is an agreed upon definition. I think most observational works would define an AGN as any detectable SMBH, i.e. any L_bol above some number. The exact number would depend on the particular survey/paper/etc. I'd suggest to just adopt a similar number as whatever observational paper you find closest to the topic you are studying. This should give a starting point.
I see. Thanks, Dylan