|Summary of Talks at Growing Black Holes 2004 in Garching
1. Ralf Bender (USM/MPE Garching) SMBH in Galaxy Centres
ca. 40 Massive Dark Objects (MDOs) confirmed today
but only for MW, NGC 4258 ( 3.9+-1.0e7) and possibly M31(9e7) SMBH nature
principal methods: stellar velocity dispersion, gas dynamics, maser dynamics, reverberation mapping, eccentric disks
bulge-less S0 galaxies : no SMBHs ?? upper limit for M33 : 1500 M_sun
anti-hierarchical formation : luminous AGN grew fast at high redshift, less luminous AGN grew fast at lower z
2. Aaron Barth (California Institute of Technology) Intermediate-mass BH in Dwarf Galaxies
Is there a lower limit for the dark halo mass below which no MBH forms ?
Problem : r_influence < 0.003’’ for sigma=40 km/sec and M_BH=e5
using AGN activity : best case NGC4395 (Sd) sigma<30 km/sec and M_BH
Greene and Ho (astro-ph/0404110) : SDSS approach
z<0.35 (Halpha still detectable)
3200 Halpha broad emission line objects (all real Seyfert1)
19 galaxies with M_BH < e6
problems : sigma(SDSS)=70 km/sec only !
LLAGN searches favour most active members of group
preliminary results : 5/7 objects with sigma<60 km/sec fit onto M-sigma relation
3. Stefanie Komossa (MPE Garching) Observational Evidence for Stellar Tidal Disruption Events and SMBH formation
Capture and disruption of Stars as mode of growing black holes (compared to accretion and BH-BH mergers)
observational evidence in NON-ACTIVE galaxies conspicuous X-ray flares (4 also known in AGN, but nature not clear)
NGC 5905 : soft X-ray flare of almost AGN luminosity, rapidly fading over timescale of months
RXJ 1242-1119 : point-like X-ray flare with an excess of a factor of 90 over the B-band lum for the host, still fading and is down by x1500 from the peak
common light curve with L = t^-5/3 ??
4. Guinevere Kauffman (MPA Garching) Accretion onto BH in the Local Universe (SDSS view)
SDSS work on BH in the local universe :122,808 galaxies (14.5 < r < 17.77) from the ‘main’ galaxy sample
Blanton et al 2003 show that the SDSS galaxy population is bimodal in color-color space, clustering properties and structural parameters such as surface brightness vs. mass. Also in stellar age as f(mass, structure), using the 4000A break and H delta indices. Division represents blue and (old) red galaxies.
In physical terms, the break is at 3e10 solar masses in stars, circular velocities of 120 km/s, concentration C=R90/R50=2.6
In SDSS, they searched for weak emission features through a careful fitting and removal of the stellar absorption features (which also yielded info. on the stellar population/M-L-ratio and age properties). Thus they could infer the star formation history.
Once they obtained emission-line strengths, they used the BPT diagram to classify normal galaxies, LINERs, and AGNs. They had to allow for the contribution of the stellar population, which they did by using [OIII] luminosities as a proxy of AGN luminosity.
Statistics : Overall fraction of AGN in the sample >50% for nearby L* galaxies at low redshift. Fraction declines toward higher z (because of increasing stellar contamination in the 3 arcsec fiber?)
The AGNs are similar to the red galaxies in terms of structural properties. In terms of stellar populations, the weak AGNs are like the old red galaxies, while the strong AGNs have younger stellar populations. The hosts represent the high mass tail and high concentration C (see above) tail of the population.
Transformation to more physical quantities by assuming L([OIII]) to L(Bol) using relation for Type 1 AGN, allowing for the unified model and star formation in the galaxy. They use D4000/Hdelta to estimate a SFR from relations for normal galaxies. Sigma-M_BH from Tremaine et al.
Finding : lower mass black holes are the ones that are more active at the present time. Characteristic BH mass overall in their sample is 1e8 solar masses. Furthermore, the low mass BHs are growing faster at the present time (because they are more active). (DOWNSIZING of BH growth)
Volume averaged star formation rate is 1000 times the accretion rate on BHs. This implies that even at the present day, SFR and activity rate are linked. The growth times of both activities track each other closely. (meaning : also SFR has down-sized ?).
For low mass BHs (1e7) half of the accreted mass comes from objects radiating within factor of 3 of the Eddington limit. However, such objects are rare, 0.1% of the total BH population Implication : bright phase/duty cycle lasts about 1e7 years.
Large-scale structure. Star-forming galaxies occur in lower-density regions. Similar results for the AGNs. Stronger activity in lower-density regions. Thus the powerful AGN fraction is higher in low density environments.
- AGN live mostly in galaxies with M>1e10 solar masses.
AGN are in galaxies with structure similar to early-type galaxies.
LLAGN have stellar populations s similar to normal early-type. High-lum. AGNs have much younger stellar populations and significant fraction have recent starbursts.
LLAGN are more strongly clustered than high-lum AGNs.
5. Smita : Black hole growth and BH mass – bulge relations of AGNs.
She starts with the BH – bulge relations of vel. disp. and mass. How did this arise and does the relation for normal galaxies apply to AGNs?
How do NLS1s relate to the normal AGN population? They seem to fall well below the normal BH mass – sigma relation. Why ?
Do the BH masses for normal AGNs and NLS1s follow the same relations? Apparently so.
How about the use of [OIII] as a surrogate? It seems to give similar results for BLS1s and NLS1s. One does have to worry about the FeII subtraction and [OIII] asymmetries in NLS1s, but they seem not to affect the mass estimates.
Thus narrow and broad-line objects do occupy distinct regions in the M_BH – sigma plane.
She finds that BH growth occurs during accretion phase in well-formed bulges. The accretion rate starts high and declines with time. AGN approach the normal BH mass – sigma at the end of the active phase. Some NLS1s lie close to the standard M sigma relation.
Separately, with Rik Williams work, they can divide the NLS1 objects according to accretion rate, and some of them (the ones near the standard M-sigma relation) have low accretion rates. Conversely, the NLS1s with high accretion rates lie below the standard M-sigma relation.
This should all be checked with more direct estimates of sigma, e.g., from the CaII triplet, CO bandhead, or the fundamental plane relations. Also, should find the locus of NLS1s on the mass-bulge luminosity plane.
Niel asks if low and high accretion NLS1s are different at E < 1keV. She says yes (low sigma NLS! Have lower alpha_x ??) Are there differences in their environments? Not known, yet.
(Then Julian gets into it. Trumper correctly suggests they discuss the question over lunch.)
6. Thomas Boller (MPE Garching) Measuring masses and accretion rates in rapidly growing young NLS1s
Extension of Tanaka’s work.
Typically, NLS1s accrete at super-Eddington rates. NLS1s have significant soft X-ray excess at low H_beta FWHM. This is reminiscent of what happens in XRBs. He says that XMM spectra of 1H0707-495 and IRAS13224-3809 have both soft X-ray excess and a steep drop at higher energies than 7keV (due to neutral Fe absorption?). But no Fe Kalpha emission.
Partial covering factor? Further observations showed that the 7 keV drop moved to higher energies with time in the former objects and occurred at 8 keV in the latter object. This could be due to high velocity outflow or ionization effects.
Their estimates show that NLS1s have moderate super-Eddington accretion rates (10-20).
Model of evolution for such a situation :
FWHM of the emission lines should increase significantly.
Furthermore, the ionizing continuum decreases (?) (thus leading to the soft X-ray excess, (which should decline with time?).
Then the soft and hard power-law indices should decrease with time.
Furthermore, FeII emission should decrease with increasing time.
Simple picture : Assume all galaxies go through an AGN phase. Start with a 2e6 solar mass BH. They find that a 0.1 solar mass/yr accretion rate will lead to a broad emission line (8000 km/s) in about 1e8 yrs, thus turning NLS1s into BLS1s. This would also lead to high-velocity outflows. Such high accretion rates imply that they live in gas-rich environments.
Note that a constraint for this approach is that the results not violate the X-ray background. We do not know the accretion-rate history of NLS1s, which is also subject to the same constraint.
Suggests we should improve the NLS1 definition by basing it on physical parameters (e.g. normalized accretion rate, m/m_dot)
Q : Do all galaxies start with NLS1 phases ?