1. How Compact are Blazar Cores – AGN studied with micro- arcsecond/sub-pc scale resolution T.P.Krichbaum with contributions from U. Bach, S. Lee, M. Gurwell and A. Marscher Max-Planck-Institut für Radioastronomie, Bonn, Germany e-mail:tkrichbaum@mpifr-bonn.mpg.de

2. involved scientists: IRAM: M. Bremer, P. Cox, S. Sanchez, C. Thum, H. Ungerechts, H. Wiesemeyer, et al. MPIfR: I. Agudo, W. Alef, U. Bach, D. Graham, T. Krichbaum, S.S. Lee, A. Lobanov, R. Porcas, A. Witzel, A. Zensus, et al. Onsala/Sest: R. Booth, J. Conway, M. Lindquist, et al. Metsähovi: A. Mujunen, M. Tornikoski, E. Valtaoja, K. Wiik, et al. NRAO-VLBA: V. Dhawan, J. Ulvestad, C. Walker, et al. and for VLBI at 1 & 2mm: ARO (HHT/KP): A. Apponi, H. Fagg, R. Freund, P. Strittmatter, L. Ziurys, et al. MIT-Haystack: S. Doeleman, A. Rogers, A. Whitney, et al. special thanks for providing data, partly prior to publication: A.Marscher & S. Jorstad (Boston) J.L. Gomez (IAA Spain) M. Gurvell (SMA) M. & H. Aller (UMRAO) T. Savolainen (Tuorla, now MPIfR)

3. What does VLBI at short millimeter wavelengths offer ? Study compact galactic and extragalactic radio sources with an angular resolution of a few ten micro-arcseconds (size, structure, kinematics, polarization) Image regions which are (self-) absorbed and therefore not observable at longer wavelength (spectrum, radiation/ energy transport, outburst – ejection relations). For nearby super-massive Black Holes a chance to study their immediate environment (Sgr A*, M87, etc.) with a spatial resolution of ~10 - 100 gravitational radii (accretion, jet formation and collimation, GR-effects).

4. Spectral variability of 3C454.3 after May 2005 Flare: Effelsberg: 1.4 – 32 GHz Pico Veleta: 90 – 230 GHz SMA: 230, 350 GHz combined data: Krichbaum, Ungerechts, Wiesemeyer, Gurwell et al. SMA data: M. Gurwell et al. quasi periodicity during peak ! 11days or 1.2  as or 95 R s T. Krichbaum, et al. 2007 monthly 2004 - 2007

5. Europe: Effelsberg (100m), Pico Veleta (30m), Plateau de Bure (35m), Onsala (20m), Metsähovi (14m) USA: 8 x VLBA (25m) The Global Millimeter VLBI Array – VLBI Imaging at 86 GHz with ~40  as resolution http://www.mpifr-bonn.mpg.de/div/vlbi/globalmm Proposal deadlines: February 1 st, October 1 st Baseline Sensitivity in Europe: 50 – 350 mJy in US: 250 – 350 mJy transatlantic: 80 – 350 mJy Array: 2 – 4 mJy / hr (assume 7 , 100sec, 512 Mbps)

6. common beam 0.25 x 0.1 mas 43 GHz, Oct. 6 86 GHz, Oct. 17 Quasi-simultaneous mm-VLBI observations of 3C454.3 after outburst 43 GHz: no emission near core known jet emission at 0.3 -1.4 mas 86 GHz: emission near core clearly visible jet components at 0.6 & 1.1mas compare well to 43 GHz image conclusion: strong absorption in the 0.1-0.3 mas region, i.e. on the 1-2 pc scale spectral index : +1.1.... + 2.6 (range of uncertainty)  43/86 GHz : - 0.4 (core), - 0.3 (jet) 8.5Jy 0.23 Jy 6.7Jy 0.14Jy 0.18Jy data: Marscher et al.

7. A new and comprehensive 86 GHz VLBI Survey: Comparison with previous surveys at 86 GHz VLBI Surveys at 86 GHz 1.Beasley et al. (1996) (N=45, 16% detect) 2.Lonsdale et al. (1998) (N=79, 14% detect) 3.Rantakyro et al. (1998) (N=67, 24% detect) 4.Lobanov et al. (2000) (N=28, 93% detect) 5.Lee et al. (2006, 2007) - 3~4 times better sensitivity (S corr >0.3 Jy) - larger sample (127sources) taken from surveys at lower frequencies - 121 (95%) sources were detected and - 109 (86%) sources could be imaged This Survey S.S. Lee et al. 2006

8. 86 GHz VLBI Survey: Some preliminary statistics distribution of core flux densities distribution of measured core sizes BL Lacs are more compact and more core dominated than QSOs. QSOs are more compact and core dominated than Radio-Galaxies. Lee, Lobanov, Krichbaum et al. 2006 Lee et al. 2007, in prep. normalized visibilities vs. uv-distance

9. M87 with the VLBA at 43 GHz Ly et al. 2004, & 2007 counter-jet or self-absorbed jet base ?

10. 234 x 59  as = 25 x 6 light days = 82 x 21 R s 9 transverse width of jet at 1mas: ~250 – 300 R s assume size of VLBI core = size of jet base: 300 R s 1 mas = 0.09 pc April 2002 96 light days

11. The Mini – Spiral VLA 2cm Krichbaum et al. 1993 VLBI 86 GHz 19 R s size Krichbaum et al. 1998 & 2006 From VLBI at 1.3mm: 50  as < FWHM < 190  as or 5 R S < size < 19 R S (for a SMBH with 4 x10 6 M  ) best estimate: 110 ± 60  as or : 11 ± 6 R S

12. 3mm VLBI of M87 with the participation of PdBure Peak: approx. 1.5 Jy

13. One year later: completely different visbility functions → the source structure must vary ! Mode: 512-8-2-LU Peak: approx 0.7 Jy

14. Variability in the inner jet of M87 detected :  0.2 mas/yr   18000 km/s (0.06c) (but: 3 – 6 c seen further downstream) 53 light days counter-jet or new jet component ? April 2002April 2003

15. new global 3mm map of M87 observed with the GMVA in April 2004: now with better sensitivity, data rate: 512 Mbit/s (MK5 hard disk recording) The size of the jet base (uniform weighting): 197 x 54  as = 21 x 6 light days = 69 x 19 R s transverse width of jet at 0.5 mas: ~174 R s M87 at 86 GHz Krichbaum et al. 2006 174 R s

16. The diameter of the light-cylinder determines the jet width: Camenzind, 1990 jet width  R s  for Kerr BH) > 40 R s Casse & Keppens 2004

17. Bach et al. 2006 and in prep. The two-sided Jets of Cygnus A Jet-to-counter-jet ratio modified by free-free absorption apparent jet speed Magnetic acceleration Vlahakis & Königl, 2003 intrinsic jet acceleration from 0.1c – 0.7c

18. Enhance 3mm VLBI sensitivity by including the 3 largest European mm-telescopes: Effelsberg 100 m (MPIfR) Pico Veleta 30 m (IRAM, Spain) Plateau de Bure, 6 x 15 m (IRAM, France) Baseline lengths (km): PdBPV EB6581700 Pdb1146 fringe spacing: 0.4 – 1.1 mas, sensitivity > 40 -70 mJy (7  512 Mbps) plus the GBT, asap : SEFD[Jy] 150GBT 590Plateau de Bure 710Pico Veleta 950Effelsberg

19. Global mm-VLBI at 150 - 230 GHz Kitt Peak, 12m HHT, 10m Metsähovi, 14m Pico Veleta, 30mAPEX, 10m angular resolutions: for 230 GHz Plateau de Bure, 6x15m near future ALMA 50x12m 20  as 60  as first detection of 2 QSOs with VLBI at 230 GHz !

20. Model of Sgr A* image distorted by gravitational light bending Reconstructed Clean Image Simulation of future 230 GHz VLBI of Sgr A* with IRAM (Pico Veleta, Plateau de Bure), HHT, CSO/SMA, CARMA, APEX

21. weighted mean:  3mm =0.18 ± 0.02 mas, range: 0.10 – 0.36 mas weighted mean:  7mm =0.55 ± 0.03 mas, range: 0.43 – 0.70 mas Does the observed 3mm size of Sgr A* vary ?

22. Conclusions The Global 3mm VLBI Array (GMVA) provides up to 40  as resolution and 50 – 250 mJy baseline sensitivity.  USE IT ! misalignement of jets between 86 GHz and longer wavelengths indicates transverse jet stratification (3C120, 3C273, 3C454.3,...). M87 and SgrA* can be imaged with nearly the same spatial resolution of a few 10 Schwarzschild radii (M87 has a jet, SgrA* not). The small diameter (~20 R S ) and high brightness temperature of the M87 jet base points towards MHD driven jet formation (Kerr metric, BH rotation acc. jet?) SgrA* is smaller than 11±6 R S ; need global 1mm VLBI (+ SMA, CARMA, HHT, APEX, ALMA,...) to image GR effects near a SMBH. Global VLBI at 2mm & 1.3mm is technical feasible. The further increase of bandwith (2-4 Gbit/s), will facilitate imaging of compact sources with a resolution of 10 - 20 micro-arcseconds at  Rest  z) ! (e.g. for z=3 and  345 GHz, Rest   Hz ). mm-VLBI monitoring is needed in support of GLAST, Herschel and Planck. The GMVA provides good sensitivity and the smallest observing beam in Astronomy.

23. StationCountry Diamete rZenith TsysGainApp.Eff.SEFD [m][K][K/Jy]%[Jy] EffelsbergGermany801300.148930 Plateau de BureFrance351200.2160570 Pico VeletaSpain301200.1455715 OnsalaSweden203000.049436100 MetsähoviFinland144000.0173023500 VLBA(8)USA251000.036202800 Hopefully soon: GBTVa, USA1001501.035150 NotoItaly321500.05203000 YebesSpain401500.2250680 NobeyamaJapan451500.1730880 Future: CARMACa,USA351500.14501070 LMTMexico501500.4360350 ALMAChile64x121001.87055 Global VLBI at 3mm: Existing and possible future antennas need more sensitivity need more stations to allow better self – calibration need southern antennas for low declinatioin sources

24. What does the GMVA offer ? a global 13 station VLBI array allowing high dynamic range imaging with angular resolution of up to 40  as at 86 GHz 3 – 4 times higher sensitivity than stand-alone VLBA (512 Mbps is standard, max. 7  baseline sensitivity is ~ 50 mJy) 2 epochs/year, each session ~ 3 – 5 days long (limitation by proposal pressure) open to cummunity by usual proposal procedures (deadlines Feb. 1 st, Oct. 1 st ) AIPS ready FITS files after correlation

25. 3C120 with the full GMVA at 86 GHz, Oct. 12, 2004 1.3 pc east-west resolution: 54 light days present sensitivity limit on long baselines ~ 0.2 – 0.3 Jy (512 Mbps)

26. quasi-simultaneous VLBI maps (within 16 days) allow spectral index determination of inner jet 3C120 at 43 GHz (data: A. Marscher et al.) misalignement possibly due to transverse opacity effects orthogonal to jet axis  jet stratification 1.3 pc 43 GHz 86 GHz -0.7 -0.3-0.4  43/86 :

27. 3C120 : Structural differences at 43 and 86 GHz within 14 days 43 GHz, Oct. 28 A B C D 86 GHz, Oct. 12 A B C D radial distances agree well, p.a.'s differ spectral index near position A: -0.4; B: 0.0; C: -1.4 (error: +/- 0.3) 0.13 pc

28. mm-VLBI resolves jet transversely: A double rail structure in the jet of 3C273 – decollimation at 3 pc ?  3.2 pc   3.5 pc  z = 0.158 1 mas  2.7 pc

29. MHD simulation of a confining B- field anchored in a rotating disk The central engine – a MHD dynamo ?

30. Signal-to-Noise ratios for the 230 GHz detections in 2003 (PV –PdB – HHT baselines): short baselines: SNR :  25 long baseline: SNR : 6 – 7 Two sources detected at 6.4 G : 3C454.3 and 0716+714 for 3C454.3 (z = 0.859) ´ = 428 GHz, life time B  2 = 3.6 10 5,   16  as = 0.1 pc = 1050 R S 9, SSA: B  1 G   600