Skip to content
A library of (semi-)empirical SED templates for type-1 AGNs
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

SED templates for type-1 AGNs and their host galaxies


  • A library of (semi-)empirical SED templates to reconcile the dust emission of type-1 AGNs with a wide range of luminosity () and redshift (), as developed in

    • Lyu, Rieke & Shi 2017 (variations of AGN intrinsic IR emission; normal bright quasars, dust-free/poor quasars)
    • Lyu & Rieke 2017 (the correct SED shape of AGN intrinsic far-IR emission)
    • Lyu & Rieke 2018 (the effects of polar dust component; low-z Seyfert-1 nuclei, extremely red quasars, AGNs with mid-IR warm-excess emission, hot dust-obscured galaxies)
  • Empirical IR templates for AGN host galaxies that have been suggested to decompose the quasar SEDs at z=5-7 (Lyu, Rieke & Alberts 2016) and those at z<4 (see Lyu & Rieke 2017)

Three flavors of type-1 AGN intrinsic IR emission
Comparison of the hot-dust-deficient (HDD) AGN template (left panel, red solid line) and the warm-dust-deficient (WDD) template (right panel, green solid line) to the normal Elvis et al. (1994) AGN template (far-IR corrected by Xu et al. 2015; blue solid line).
see Lyu, Rieke & Shi 2017 for details
Usage Demonstrations of the Reddened AGN Templates
NGC 3783: a Seyfert-1 AGN with confirmed mid-IR polar dust emission Extremely Red Quasars Hot Dust-Obscured Galaxies
see Lyu & Rieke 2018 for details
Host galaxy FIR templates Host galaxy stellar emission template
Comparison of Haro 11 template (red line) and the normal SFG templates in Rieke et al. (2009) (blue lines). Comparisons of different stellar templates. The final empirical stellar template is shown as a green line.
see Lyu, Rieke & Alberts 2016 for details see Lyu & Rieke 2018 for details

Directory information

        AGN_torus_faceon.temp.full.txt             -- intrinsic templates for normal, warm-dust-deficient and hot-dust-deficient AGNs
        AGN_polar_emission.temp.full.txt           -- template for the polar dust emission as trained for low-z Seyfert-1 nuclei

              norm-[id].dat    -- reddened templates for normal (NORM) AGNs, tau_{pol, V}= 0 - 5
              hdd-[id].dat     -- reddened templates for hot-dust-deficient (HDD) AGNs, tau_{pol, V}= 0 - 5
              wdd-[id].dat     -- reddened tempaltes for warm-dust-deficient (WDD) AGNs, tau_{pol, V}= 0 - 5
              hdo-[id].dat     -- hot dust-obscured (HDO) templates for normal AGNs, tau_{pol, V}= 0 - 5

        obsagn_temp.norm.index.lis   -- ID, tau_{pol, V} and f_{pol, 10\mum} information for templates in norm/
        obsagn_temp.hdd.index.lis    --  **** in hdd/           
        obsagn_temp.wdd.index.lis    --  **** in wdd/
        obsagn_temp.hdo.index.lis    --  **** in hdo/

              rieke-F[loglum]_temp.txt   -- Rieke + 2009 SFG tempaltes, loglum = alog10(L_{IR}/L_sun)
              Haro11.temp.full.txt       -- Haro 11 far-IR template, suggested to be used for quasar hosts at z>5
        old_stars.temp.full.txt          -- template for old stellar population with mid-IR dust features

   plots/        -- Figures for the README file.

Usage Guidelines


The AGN templates here have been constructed in a "minimalist" style: we wish to build reasonably good SED models with the least numbers of free parameters as possible. Based on real observations, we derived three intrinsic templates to describe the typical SED shapes of accretion disk + torus emission as seen from a face-on viewing angle (aka, Type-1), tested their robustness with independent calibrations available for PG quasars, and demonstrated such variations could be common at different redshifts and AGN luminosities. Then we introduced a very simple polar dust model that could obscure the intrinsic AGN templates, and showed that it worked reasonably well to reproduce the polar dust emission strength observed in nearby Seyfert-1 nuclei and to reconcile the SED behaviors of various type-1 AGN populations. These templates can be generally applied in e.g., SED decompositions of galaxies with possible AGN contribution or quasars, IR AGN Selections and the search for peculiar objects, looking for AGN candidates with strong polar dust emission, credibility check of the torus model SEDs from a radiative transfer code.

How to use them in SED decompositions?

Although these AGN templates are empirically constructed and tested against various observations, they should be applied correctly to get scientifically meaningful results (you should never trust a fitting result merely because of a very small chi-square). For example, you should not add an old stellar template to fit the SED of a z~5 quasar (the host galaxy is too young!).

For a type-1 AGN, its optical to far-IR SED can be modeled by a linear combination of various templates.

f_{tot} =  f_{AGN_temp} +  f_{host_IR}  
                                          + f{old_stars}, if z<3-4  
                                          + f_{AGN,syn},  if the AGN is radio-loud

where the AGN component

f_{AGN_temp} = any of the NORM, WDD, or HDD intrinsic AGN template (or a combination of two), if there is 
               enough reason to believe that the polar dust component is very weak or non-exist.
             = any of the reddened NORM, WDD or HDD AGN templates
             = any of the three intrinsic AGN templates + template for polar dust emission , if the goal is to 
	   get very crude estimations

	    1. For reddened type-1 AGNs, you should try NORM, then WDD, and lastly HDD. 
	       If none of them works, try HDO;
	    2. It is highly recommanded to use mid-IR spectral decompositions or high-spatial resolution
	       SED data to check if the SED results are consistent whenever possible;
	    3. You are suggested to check the optical and the IR SEDs together, especially if
	       an composite SEDs for a large sample can be built;
	    4. The polar dust could be clumpy and the galaxy ISM dust may also cause the line-of-sight
	       obscuration. As a result, it is OK if the individual UV-optical SED is not matched by the 
	       AGN template.

the host galaxy IR template

f_{host_IR} = any of the Rieke+09 templates with loglum = 9.5-12.5, if z <4-5
            = Haro 11 template, if z=5-7 optically-selected quasars

(to be completed)

How to infer the polar dust emission strength?

Before any fittings, the host galaxy mid-IR contribution should be properly removed (or considered in the SED decompositions).

Find the best reddened AGN templates that matched the overall SEDs together with other templates, then look for the corresponding f{pol, s+e} value in obsagn_temp.[type].index.lis. f{pol, s+e} is the relative contribution of the scattering (s) and emission (e) components of the polar dust at 10 microns (f{pol, s+e}= f{pol, s} + f{pol,e}).

You may also combine any one of the three intrinsic AGN templates with the template for the polar dust emission to get an estimation.

(to be completed)

Is there a fitting code for these templates that can be used directly?

Yes, it should be avaliable here by the end of 2018, both in IDL and python versions. I'm still doing some tests and cleaning the code. Stay tuned!

Cautionary Notes

  • All the AGN and galaxy templates have been interpolated on the same wavelength grids with relatively high spectral resolutions (0.01-1000 microns with 5000 points in log scale). However, these templates are constructed to reproduce the broad-band SEDs only;

  • The AGN templates are designed to be used only for type-1 objects (aka. AGNs with broad emission lines). We currently do not have a good idea on the IR emission behaviors of the "edge-on" torus, given the possible contaminations of polar dust and host galaxy dust;

  • The tau_V associated with the reddened AGN template describes the integrated optical depth of the polar dust component, not necessary the value along the line-of-sight. tau_v is equal to the radial optical depth of a spherical shell if the dust grains in the polar component at the same radius are evenly redistributed on the same sphere;

(to be completed)


Please contact me via jianwei (at) if you have any comments, questions or suggestions.

You can’t perform that action at this time.