1. Comment on two topics related to fundamental “exotic” physics with AGN Alessandro De Angelis Brera, March 2012
Testing fundamental spacetime symmetries
AGN spectra and photon/ALP oscillation

2. Rapid variability and LIV
MAGIC, Mkn 501
Doubling time ~ 2 min
astro-ph/
arXiv:
HESS PKS 2155
z = 0.116
July 2006
Peak flux ~15 x Crab
~50 x average
Doubling times
1-3 min
RBH/c ~ s
H.E.S.S.
arXiV:

3. GRBs Another probe
Interesting for astrophysical reasons, for propagation physics, for rapid variability
Fermi is changing our view, due to its unprecedented range, self-pointing, dedicated instrument
MAGIC is the best IACT, due to its fast movement & low threshold
De Angelis, Galanti, Roncadelli 2011
with Franceschini EBL modeling
MAGIC
H.E.S.S.
redshift z
g ray energy (TeV)
t = 1 (GRH)
> 1
region of opacity
No VHE g emission from GRB positively detected
yet...
(all observed GRB very
short or at very high z)
But how many can
we expect/year?
0.02 < N < 0.5

4. Mkn501 flare: violation of the Lorentz Invariance
Mkn501 flare: violation of the Lorentz Invariance? Light dispersion expected in some QG models, but interesting “per-se”
V = c [1 +- x (E/Es1) +– x2 (E/Es2)2 +- …]
1st order
TeV
TeV
TeV
TeV
> 1 GeV
< 5 MeV
4 min lag

5. LIV in Fermi vs. MAGIC+HESS
GRB080916C at z~4.2 : 13.2 GeV photon detected by Fermi 16.5 s after GBM trigger. At 1st order
The MAGIC result for Mkn501 at z= is Dt = ( ) s/GeV; for HESS at z~0.116, according to Ellis et al., Feb 09, Dt = ( ) s/GeV
Dt ~ (0.43 ± 0.19) K(z) s/GeV
Extrapolating, you get from Fermi ( ) s (J. Ellis et al., Feb 2009)
SURPRISINGLY CONSISTENT:
DIFFERENT ENERGY RANGE
DIFFERENT DISTANCE ~z

6. Fermi: GRB GRB
z = ± 0.003
prompt spectrum detected, significant deviation from Band function at high E
High energy photon detected:
31 GeV at To s
[expected from Ellis & al. (12 ± 5) s]
tight constraint on Lorentz Invariance Violation:
Es1 = MQG > several MPlanck
z = 1.8 ± 0.4
one of the brightest GRBs
observed by LAT
after prompt phase, power-low emission persists in the LAT data as late as 1 ks post trigger:
highest E photon so far detected: 33.4 GeV, 82 s after GBM trigger
[expected from Ellis & al. (26 ± 13) s]
much weaker constraints on LIV Es
=> Fermi rules, but 1st order violations appear unlikely

7. 2nd order? Cherenkov rules!
Es2 > GeV (10-9 MP) (HESS, MAGIC)
Difficult to improve by 9 orders of magnitude 

8. Interpretation of the results on rapid variability
The most likely interpretation is that delays are due to physics at the source
However
We are sensitive to linear effects at the Planck mass scale
2nd order effects: the realm of Cherenkov detectors
Anisotropy?
More observations of flares might clarify the situation (or complicate it)

9. Propagation of g-rays e+ e- gVHEgbck  e+e- Mean free pathx
gVHEgbck  e+e-
dominant process for absorption:
maximal for:
Heitler 1960
s(b) ~ ≈
For gamma rays, relevant background component is optical/infrared (EBL)
different models for EBL: minimum density given by cosmology/star formation
(Galanti et al. 2012)
Mean free path
Measurement of spectral features permits to constrain EBL models
(Dominguez et al. 2011)

10. Are our AGN observations consistent with theory?
Selection bias?
New physics ?
Measured spectra affected by attenuation in the EBL:
observed spectral index
~ E-2
redshift
(DA, Galanti, Roncadelli; PRD 2011)
The most distant: MAGIC 3C 279 (z=0.54)
Alessandro De Angelis

11. Attempts to quantify the problem overall
Analysis of AGN
For each data point, a corresponding lower limit on the optical depth t is calculated using the minimum EBL model of Kneiske & Dole 2010
Nonparametric test of consistency
Disagreement with data: overall significance of 4.2 s
=> Understand experimentally the outliers
(Horns , Meyer 2011)

12. If there is a problem Explanations from the standard ones
very hard emission mechanisms with intrinsic slope < 1.5 (Stecker 2008)
Very low EBL, plus observational bias
to almost standard
gamma-ray fluxes enhanced by relatively nearby production by interactions of primary cosmic rays or n emitted from the same source
to possible evidence for new physics
Oscillation to a light “axion”? (DA, Roncadelli & MAnsutti [DARMA], PLB2008, PRD2007, PRD2011)
Axion emission (Hooper et al., PRD2008)
A combination of the above (Sanchez Conde et al.)

13. Recent confirmation

14. PKS1222: an alternative detailed discussion of the problem
Tavecchio, Roncadelli, Galanti, Bonnoli 2012:
the g → a conversion occurs before most of the photons reach the BLR. Accordingly, ALPs traverse this region unimpeded and
Outside, the re-conversion a → g takes place either in the same magnetic field of the source or in that of the host galaxy
Resulting parameters of the ALP needed to fit the data consistent with DARMA

15. Summarizing: if the expected photon yield at VHE is different from what we think, what might be wrong?
Emission models are more complicated than we think
VHE photons are generated on the way
Something is wrong in the gg -> e+e- rate calculation
gg -> e+e- cross section
Boost (relativity)
QED
(2nd order effects due to vacuum energy could not decrease the absorpyion probability)
Vacuum energy (new sterile particles coupling to the photons)
“Se non e’ vero e’ ben pensato”: consistent values for m, (1/M) in a range not experimentally excluded
But explanations are expensive: an ALP

16. We are (maybe) making two extraordinary claims
A possible relation between arrival time and energy
Signal from sources far away hardly compatible with EBL
But warning: how hardly?
We should keep in mind that
Extraordinary claims require extraordinary evidence
New Scientist, SciAm blog/news, …, and then?
Claims must be followed up
If we see this in such sources, what else do we expect?
Fundamental implications of unexpected findings?
Are we seeing a part of the same big picture?
Warning: this is a very connected world… If you touch
dispersion relations, the Universe might become transparent
to photons beyond the GZK (Galaverni & Sigl 2008)

17. ce = cg(1+d), 0 < abs(d) << 1 Coleman & Glashow; Stecker and Glashow
If d<0 => ce < cg => decay g -> e+e- kinematically allowed for gamma with energies above
Emax = me sqrt(2/abs(d))
- Eg > 100 TeV => abs(d) < 5 x 10-17
2. If d>0 => ce > cg => electrons become superluminal for energies larger than Emax/sqrt(2) => Vacuum Cherenkov Radiation.
- Ee > 2 TeV from cosmic electron radiation => abs(d) < 2 x 10-14
Modification of g g -> e+e- threshold. Using Mkn 501 and Mkn 421 spectra observations up to Eg > 20 TeV
=> abs(d) < 1.3 x 10-15
From MAGIC Mkn501 (taken as a LIV signal): |d| ~

18. A no-loss situation: if propagation is standard, cosmology with AGN
GRH depends on the –ray path and there the Hubble constant and the cosmological densities enter => if EBL density and intrinsic spectra are known, the GRH might be used as a distance estimator
GRH behaves differently than other observables already used for cosmology measurements.
Blanch & Martinez 2004
Simulated
measurements
Mkn 421
Mkn 501
1ES
Mkn 180
1ES
PKS
1ES
1ES H
PKS H
EBL constraints are paving the way for the use of AGN to fit WM and WL …

19. Determination of H0, M , L
Using the foreseen precision on the GRH (distance at which t(E,z)=1) measurements of 20 extrapolated AGN at z>0.2, cosmological parameters can be fitted.
=> The Dc2= parameter contour might improve by a factor 2 the 2004’ Supernovae combined result !

20. Conclusions
When we were young we have been happy: we found possible hints of new physics in 2 of the most promising AGN channels
However, it looks to me that we are a bit stuck now
What should we do if we don’t want to keep repeating the same talks and we believe on what we are doing?
Gamma propagation and transparency of the Universe:
Detect well some AGN at z > 0.2 (one of the goals to consider for focusing our scientific objectives)
Know better the relevant astrophysics: make the best possible computation of the VHE SED at the source, and of EBL (and of magnetic fields?)
(a no-loss situation in which we improve in the worst case our knowledge of cosmological parameters)
Gamma propagation and possible LIV
Detect more rapid flares (how?), possibly far away from us
Have reliable models on the energy profile of the emission at different times
(a no-loss in which we improve in the worst case the limits on 2nd order LIV)