(23-27 Jan 2023)
Getting ready to descend the slippery slope of multi-messenger cosmological black holes data
No notes for myself!
- (2021) SAM of BH growth
- Focusing on z = (6, 2, 0.2) to reproduce three particular QSOs, reproducing their properties and detectibility of BBHs
- 1000s of light seeds and 20-40 heavy seeds in their progenitors
- 100-1000 Msun BHs not detectable by even Lynx or Athena. Their peaks are in the EUV. Only detectable with Einstein Telescope.
- (Trinca+ 2021ab) CAT: New SAM. Only JADES-Deep will detect MBHs between 10^4 - 10^6 Msun
- (in prep) AGN+galaxy SEDs with emission lines calculated with CAT and CLOUDY
- (2021) Heavy seed formation in QSO hosts: tested Wise+(2019) rapid growth scenario in DM zoom-in simulations
- Built a SAM aimed at heavy seed formation
- A factor of 500 times more atomic cooling halos in z~6 QSO hosts than an average field
- There are about 100-500 close pairs in the overdense region (QSO progenitors)
- LW radiation allows the halo to grow without forming stars to reach the close pair / rapid growth scenarios
- For the pristine halos above the dynamical heating threshold, they have a typical LW intensity of 100 J_21. For all halos, this reaches up to 10^5 J_21.
- Massive BHs in low-mass galaxies
- (Calura+ 2022) SIEGE simulations with ~100 Msun DM resolution and single star formation / feedback. Run with RAMSES.
- (Lupi+ in prep) Using GIZMO and KROME for z~6 dwarf galaxy (2 x 10^9 Msun). Single star formation, sampling from an IMF, for both Pop II/I and III. BHs do not grow after forming between z=10-20 and their initial growth spurt.
- (Civano+ 2019) Observability of early SMBHs and their host galaxies
- (Kashlinsky+ 2012) Using the excesses in the CIB and CXB to determine any unresolved sources. Could be related with high-z BHs, but seeds (both DCBHs and Pop III remnants) cannot explain all of it. Also see Ricarte et al. (2019).
- (paper?) OGLE has detected ~10 long-duration microlensing events. Primordial BHs (PBHs)?
- (Garcia-Bellido+ 2019) Calculated PBH mass spectrum formed during QCD phase transitions. Roughly constant below ~0.1 Msun and peaking just under the Chandrasekhar mass and sharply falls off as a power law.
- (Cappelluti+ 2022) PBHs can accelerate halo growth in small minihalos (< 10^5 Msun). This can drive star and galaxy formation much more differently than LCDM, where the cSFR and ionization fraction can have a non-monotonic evolution.
- Performed MC realizations of PBHs around Sgr A* with GW radiation hardening within their model. The PBHs do not merge in their simulations but emit continuous GWs (mainly caused by harmonics?) as they orbit Sgr A*.
- LISA only has a 10% chance to detect one PBH within 10 years
- However µAres (also see this recent paper) could detect ~140 PBHs.
- In any case, these events could create some sort of GW background
- (Bondani+ in prep) Now they're considering eccentric orbits, which initial distribution follows a thermal one.
- In this improved model, LISA has a 60% chance of observing 1+ sources.
- µAres will resolve all sources in this model.
- (Ziparo+ 2022) Model considers an isothermal gas sphere with PBHs following an NFW profile.
- As the PBHs travel, they may past through halo centers, which the accretion rates will be 10^4 higher than the IGM. They also consider cosmological HII regions around halos for accretion rates.
- As the PBHs accrete, they heat their surroundings and can increase the minimum halo mass of galaxy hosts, affecting any type of radiation backgrounds.
- Can explain all of the CXB excess but 1% of the cosmic radio background excess.
- L-Galaxies is the Munich galaxy SAM. New work here is a 30 Mpc cosmological volume that can resolve 10^6 Msun minihalos to address SMBH seeding. Named "Millennium Seeds"
- Using GQd (see Valiante talk on Monday) for Pop III seed "grafting" into the merger trees. Also considers DCBH and IMBH formation from the LW intensity and metal enrichment.
- (Spinoso+ 2022) DCBH/IMBH hosts are generally in more dense environment when compared to Pop III remnants.
- (Mezcua+ 2023) Investigated z~1 AGN and found that stellar masses are comparable to the SMBH masses. Similar to overly massive BHs at high-z?
- (Izquierdo-Villalba+ 2020) Investigated wandering BHs, either in isolation or recoiling. The SMBHs in pseudobulges are generally below the M-sigma relation.
- Spin is important because it affects feedback and recoil (if there's a merger). Basically the ISCO is a function of spin.
- In a retrograde thin disk, the radiative efficiency is a few percent. For a non-spinning disk, the efficiency is ~10%, and this goes up to 40% with a prograde disk.
- For a thick disk and jets, the efficiency is increased by a factor of ~3. The spin evolution (da/dt) is dependent on the accretion disk structure and the angular momentum at the ISCO (e.g. can spin-down a BH in a prograde disk).
- Introduced an on-the-fly spin evolution model in the NewHorizons simulation (16 Mpc zoom-in region with 34 pc resolution).
- Model has a jet mode (thick disk; f_Edd < 0.01), quasar mode (thin disk; f_Edd = 0.01-1), and super-Eddington mode.
- Sees varied spin histories, depending on the types of feedback modes the SMBHs have experienced. Most have spins >=0.5. They spin up very quickly after formation.
- Nearly all maximally spinning BHs (a > 0.9) are in the quasar mode (no jet spindown; gas rich?).
- (Massoneau+ 2022) Looked at super-Eddington accretion, which delays but doesn't prevent spin-up. As the BH spins up, super-Eddington accretion becomes rarer, so the long-term impact on mass evolution is small.
- Obelisk simulations: BH merger dynamical delays. Two BH particles merge when they are closer than 140 pc.
- In post-processing, the merger is delayed by a dynamical friction timescle (with stars) and binary hardening phase (by stars or viscous torques in a circumbinary disk).
- Low-mass (Mstar < 10^9 Msun) galaxies have chaotic dynamics and no well-defined center -> chaotic accretion
- Larger galaxies have a disk, causing coherent and efficient accretion.
- Above 10^11 Msun, the galaxies become gas-poor and accretion rates decrease. Mergers drive growth.
- The delayed merger model shifts the BH masses at merger to higher values by an order of magnitude (log M = 5-6
$\rightarrow$ 6-7) - Merging MBHs tend to have higher spins than the overall population, and then mergers decrease the remnant.
- For galaxies hosting mergers, the sSFRs and BH accretion rates are generally higher in the mergers in the simulation. However in the delayed model, the gas-rich environment dissipates and the sSFR boost is less and a smaller f_Edd boost.
- Merger remnants are strongly outshined by the galaxy in the UV by ~5 magnitudes. However, there are some hosts that are AGN dominated and can be detected in X-rays.
- The detectibility in X-rays and not UV/optical biases observations toward efficiently radiating BHs at the high-mass (>10^7 Msun) end, missing slowly accreting BHs and/or lower-mass ones.
- About 5% are detectable in the radio by SKA and ngVLA.
- Nice BBH orbital evolution plot from LISA white paper (Bortolas) with all of the relevant processes
- Potential bottlenecks for kpc-scale:
- Delayed BBH pairing in minor mergers (e.g. Tremmel+ 2018)
- Global non-axisymmetric gravitation torques versus dynamical friction (e.g. Bortolas+ 2020, 2022)
- Clumps in galactic and circumnuclear disks
- Effects of stellar + AGN feedback (e.g. Gruzinov+ 2020, Bellovary+ 2021, Roskar+ 2015)
- Low-mass galaxies: sensitivity on the inner DM profile (e.g. Tamfal+ 2018)
- (Mayer+ 2016) typical stellar mass of clumps(?) are between
$10^6 - 10^8 M_\odot$ , regardless of simulation code. - (Dessauges-Zavadsky+ 2023) ALMA CO observations have detected oversized GMC that have similar masses to the above study.
- (Bortolas+ 2020) Investigated the MBH sinking time in a gas-rich z=7 galaxy. Found that gravitational torques were dominate over the dynamical friction.
- (Rizzuto+ 2021, 2022) Studied direct N-body simulations of dense stellar clusters. They also looked at the frequency of both TDEs and tidal capture events. About 500-1000 events over 40 Myr of evolution.
- Found that 25% of the stars within the sphere of influence of the MBHs, they form a bound Bahcall-Wolf system from which they can undergo tidal events, which can cause BH growth.
- (Partmann+ in prep) Looking into sinking and merging of (
$10^3 - 10^7 M_\odot$ ) IMBHs within dwarf galaxies ($2 \times 10^9 M_\odot$ ). 5:1 mergers. Many different outcomes: no sinking, 3-body ejecta, Gw recoil ejection - When a massive BH is ejected, the DM inner profile becomes cored, similar to a gas blowout by supernova. Rantala+ (2018, 2019) has found a similar coring process caused by the orbiting and sinking BH.
- (Frigo+ 2021) Core formation occurs during the halo merger, while the velocity anisotropy varies
- (Franchini+ 2022) Using GIZMO simulations of a circumbinary disk (CBD), they find that gravitational torques come from circumstellar disks (r = 0-a), streams (r = a-2a), and the circumbinary disk (r > 2a).
- They explore different initial disk temperatures, which is effectively the h/r ratio (Farris+ 2014). This affects the dynamics with the warmer disk having less structure.
- (in prep) They're now exploring a "live" BBH orbit versus a fixed orbit. This marginally increases the accretion torque (only in the minidisks) and removes any type of accretion rate modulation w.r.t. the orbital period.
- (Franchini+ 2021) CBD Simulations with a self-gravitating disk, which is unstable and forms spiral arms. The binary shrinkage increases when compared to without self-gravity.
- Several phases of EM emission during the merger. On top of the binary system and CBD, there are the minidisks, jets, and recoil to consider and how it interacts with the ambient medium.
- (Lops, Izquierdo+ 2023) Event estimates with LISA and X-ray pointings
- Overall GW signal: need population parameters (galaxy merger rates, MBH masses) and local dynamics (accretion and BBH environment)
- (Middleton+ 2021) PTA signal interpretation.
- There has been a claimed PTA signal. It could be caused by several different types of sources (e.g. cosmic strings), but it could be explained by massive BBHs.
- Assuming that BBHs are the sources, some constraints can be placed: merger timescale is < 3 Gyr and that they come from galaxies/SMBHs on the high-end of the M-sigma relation.
- (Izquierdo-Villalba+ 2022) The BBHs responsible for the PTA signal overproduce the QSO luminosity function, so this suggests that a fraction should be obscured.
- (Volonteri+ 2017, 2022) SEDs of AGN+galaxies, using a SS disk and X-ray power-law but no emission lines
- (Zhang+ 2023 at U. Arizona) With the TRINTY code, they devise an empirical model for MBH populations out to z=10. Predicts smaller MBHs in low-mass galaxies at higher redshifts.
- (Volonetri+ 2022) Taking the JWST early results to search for AGN signatures.
- The AGN signatures are much fainter than the galaxies in nearly all JWST bands for even overmassive (M_BH / Mstar = 0.01-0.1) MBHs.
- They also explore their placement in color-color diagrams to see how they dependent on M_BH/Mstar ratios.
- Only overmassive BHs can be selected in this manner, which will give you a biased selection of AGN at high-z.
- (Massonneau+ 2022) Simulations of super-Eddington accretion and feedback
- They find that the average growth rates are less than Eddington accretion happening at all times
- Only with weak jet/wind feedback, the average accretion rates are comparable to Eddington accretion rate.
- (Sassano+ 2021) Pathways to BH seeds, depending on LW radiation, dust cooling, and metallicity.
- (Chon & Omukai 2020) Competitive gas accretion can also create medium/high-mass BH seeds
- (Trinca+ 2022) Evolution of nuclear BH mass function, depending on seeding models and accretion (i.e. Eddington vs super-Eddington) models.
- (Sassano+ 2023) GASOLINE2 simulations of super-Eddington on an IMBH in a first galaxy's nuclear disk before any SF occurs.
- After the first 1 Myr of rapid growth, there is limited growth because the disk fragments
- The accretion rate is very stochastic because of the accretion of the clumps
- Using these simulation results, they predict the growth from z=15 to z=6 by a factor of ~1000.
- Typical models have
$M_{BH} = 5 \times 10^6 M_\odot$ and$M_\star = 10^9 M_\odot$
- (Maiolino+ in prep; Valiante+ in prep) The observed JWST AGN candidates are generally above the z=0 M-sigma relation. Also in a BPT diagram, they would not be classified as AGN (see Nakajima & Maiolino 2022) because of a low-metallicity environment.
- One method to identify Unresolved BHB candidates: peculiar spectral properties
$\rightarrow$ selects loose binaries.- In this case, the broad-line region is only attached to one BH.
- Observe the broad lines' velocity shifts with respect to the narrow lines.
- However, the orbital periods can be up to 100 years.
- Searches: Tsalmantza+ (2011), Eracleous+ (2012)
- Alternative explanations for shifts: natural varability, recoiling MBHs, disky BLR, cosmological (unlikely) / galaxy cluster superposition
- A novel and faster binarity test: Use a variation of reverberation mapping
- Cross-correlation the red and blue sides of the line profile
- In a mock equal-mass binary, there is an anti-correlation at the line center instead of a maximal correlation in the case of a single BH. Works even for wide binaries up to 1000 year orbital periods.
- However, there are very few systems that this method is applicable.
- (Westernacher-Schneider+ 2022) Circumbinary simulation: 2D GPU-accelerated code Sailfish.
- Explored parameters: eccentricity, mass ratio, and disk temperature
- Optical/IR variability is cleaner in eccentric orbits.
- q = 0.3-1: Bursty variability on the orbital time on the cavity wall
- q = 0.05-0.3: More sinusoidal varability
- (Xin & Haiman 2021) Searching for binaries in LSST
- There should be O(100) binaries with a period <1 day in LSST at z=1-2 and log M = 5-6.
- If a binary is detected in LISA, then one can search for the source in the LSST archives.
- (Davelaar & Haiman 2022ab) Binary self-lensing
- There are recurring flux spikes if the binary is nearly edge-on (+/- 30 degrees).
- There is a dip in the spike corresponding to the emission hole. To detect needs to be edge-on within 10 degrees.
- Should be O(100) binaries in LSST that are aligned.
- One candidate: Hu, D'Orazio, Haiman+ (2020) well fit by an eccentric binary, total mass =
$3 \times 10^7 M_\odot$ , q = 0.2, T = 418 days, e = 0.5, inclination = 8 deg
- (Krauth+ in prep) Follows last month (400 orbits) before merge with 2D simulations with Sailfish.
- The UV/X-ray luminosity is nearly constant (with decreasing periodicity) until the last 2 days.
- X-rays drop by 5 orders of magnitude due to the disappearance of the minidisks.
- Using L-Galaxies on the Millenium Simulation to model the MBH binary population.
- Create a light cone up to z = 4 to search for single and binary SMBHs and ones in an active phase (i.e. AGN).
- For
$M_{BH} < 10^9 M_\odot$ , 75% are hosted by spirals with bulges - For larger SMBHs, they're hosted in spirals at z > 1, but are hosted by ellipticals at later times.
- For LISA EM follow-ups at z = 2-3, within the sky localizaiton error box, there are 10^6 galaxies at 10 hours before merger, decreases to 10^5 galaxies 1 hour before, and 1000 galaxies at merger.
- However these numbers decrease by 4 orders of magnitude at z < 0.5.
- Can use X-ray emission to reduce the host galaxy candidates by a factor of ~30.
- Use TDEs, EMRIs, and massive binaries to explore the (I)MBH mass function with future observatories (LSST, ULTRASAT, LISA, eROSITA, etc)
- TDE expected (observed) rate ~
$3 \times 10^{-4} (10^{-5})$ per galaxy per year - (Miles+ 2020) Partial TDE light curves and accretion rates. Candidates: Payne+ (2021), Liu+ (2023)
- (Ryu+ 2020ab) GRHD simulations of partial TDEs to find critical radii. If a star comes within 10 tidal disruption radii, partial TDE will occur. Provides fits as a function of BH mass and mass stripping fraction.
- (in prep) Standard estimates for TDEs may be overestimated by 10-100 times in nucleated galaxies wtih an empty loss cone, which can reconcile the discrepancy between the observed and expected rates.
- (in prep) If we define the loss cone with the partial (
$\Delta m / m > 3$ %) instead of complete TDE radius, this increases the expected TDE rate by a factor of 50.- Very eccentric orbits are easily deflected out of the loss cone.
- Less eccentric orbits relations is inefficient within an orbital period.
- We should define a radius for classifying a partial TDE by connecting it with observables.
- (Bonetti+) Considering EMRI formation in a MBH binary, they find a rate 10-100 times larger (
$10^{-5} - 10^{-6} M_\odot ; {\rm yr}^{-1}$ ) than the standard 2-body relaxation channel around a single MBH. - Also, there is a formation burst, where the orbital period sets the lifetime of the burst.
- Coupling with the galaxy SAM, L-Galaxies, they predict 60-75 EMRIs per year in the full sky. But only 10-15% will be detectable by LISA with SNR > 20.
- (Andalman+ 2022; Steinberg & Stone 2023) Moving mesh simulations of TDEs. Highly non-ideal, unlikely previous simulations having Keplerian orbits.
- (in prep) Using L-Galaxies with BHs (see Bonoli's and Izquierdo's talk) and Millenium II
- Modifying the above model to include a nuclear stellar cluster, phenomenologically.
- For MBHs with
$\le 10^6 M_\odot$ , its inclusion changes the expected event rate by making it decrease with time instead of a constant rate. - TDEs can account for 75% of accretion at z>6 but <5% at z=0 for a
$\le 10^6 M_\odot$ BH.
- If there's any gaseous effects on the inspiral, it'll cause a dephasing from the expected waveform. Sensitive to local disk parameters.
- (Garg+ 2022) High-z BH seed growth requires very strong accretion. Any type of dephasing may constrain accretion flows.
- (Zwick+ 2022) Inhomogeneities within the disk will cause variability, causing a "dirty waveform".
- These produce a background possibly detected with LISA
- The SNR depends on the turbulent spectrum, e.g. Kolmogorov
$\rightarrow$ SNR = 30
- OJ287 is a flaring QSO with a "long-short" modulation (e.g. Dey+ 2019). 2.5 PN dynamics gives 10% accuracy. 3.5 PN can give you the BH spin.
- The flares are possibly caused by a secondary BH passing through the primary's disk.
- The flare's properties suggest a gas density of
$10^{15} ; {\rm cm}^{-3}$ .
- "Changing look" AGN: these dramatically change their classification and spectral appearance (e.g. LaMassa+ 2015; Runnoe+ 2015 are the first ones detected).
- Review: Ricci & Trakhtenbrot (2022, in press)
- The latter paper saw the QSO spectral component (lines and continuum) completely disappear from 2003 to 2015.
- (MacLeod+ 2016) Another example. Overall there are about 2 dozen of them.
- Are these changes in accretion or obscuration?
- For obscuration to work, it needs to hide both the entire BLR and accretion disk.
- To test whether accretion is the cause, observations need to follow the sequence of the disk brightening and then 10-100 days later, the BLR changing while the NLR remains unchanged.
- They detected such a system (1ES 1927+654), where the BLR appeared from nothing (before only NLR) 3 months after the continuum flare.
- What caused the flare in the first place?
- A TDE? (Merloni+ 2015; Chan+ 2019, 2020)
- B-field inversion event? (Scepi+ 2021; Laha+ 2022)
- (Zeltyn+ 2022) Example is explained by variable obscuration, not accretion.
- Implied cloud velocity ~ 0.5c
- Dimming about a year with the recovery to the bright state within 2 months.
- eROSITA has a large effective area and does all-sky surveys every 6 months due to celestial mechanics (non-pointing telescope). Early data release in 2021.
- eFEDS: 1% preview survey. 140 deg^2, 2.5 ks exposure, 27k X-ray sources, 12k spec-z, flux limit ~ 7e-15 erg/s/cm^2.
- (Liu+ 2023) Star on a tightly bound orbit around a SMBH? Repeating X-Ray/UV/radio flares; potential change in recurrence time due to decrease in orbital period(?)
- Major discoveries: hot phase of the MW and local group; statistical study of CGM; AGN feedback constraints, etc.