Unidentified gray sources at high Galactic latitudes

1
Unidentified g-ray sources at high Galactic
latitudes

• Diego F. Torres
• dtorres_at_igpp.ucllnl.org

www.angelfire.com/id/dtorres
2
Outline

• Low, Mid, and High Latitude sources
• High Latitude gamma-ray fluxes
• AGNs
• Comments on phenomenology, theoretical models,
  lensing, prospects
• Are AGNs all there is?
• Open questions for diffuse emission contributions
  and galaxy clusters (to be covered in other
  talks)
• Alternative plausible sources
• Radiogalaxies
• Normal Galaxies, Starbursts, LIRGs
• High latitude molecular clouds

3
Low-, Mid-, and High-Latitude Sources different
populations
Gehrels et al. 2000

• Finding
• excess of weak sources at mid-latitudes
• lack of high-latitude strong sources.

4
Possible origin for the mid-latitude sourcesThe
Gould Belt
Grenier, 1995, 2000
Intrinsically weaker sources Gould Belt
distances typically less than 500 pc (as compared
to several kpc of the low latitude detections).
The Gould Belt is an starburst disk (massive
stars are represented as white dots) that
dominates our neighborhood up to a thousand light
years away, and includes many of the brightest
stars in the sky. Many of the stable,
unidentified gamma-ray sources found by EGRET at
mid-latitudes (yellow dots) closely follow the
curve and apparent width of the belt.
5
Log N Log S at high latitudesdistinguishing
peak and average fluxes
Reimer Thomson 2001

• Plausibly identified AGNs, and UNIDs show
  significantly different peak and average fluxes
• Sources are preferentially identified by peak
  fluxes only.
• Similarity between peak flux distributions
  support physical association between UNIDs and
  AGNs.

6
Active Galactic Nuclei

• Supermassive BH accreting matter from accretion
  disk.
• Jets of relativistic particles emanating from the
  center.
• Unification of different classes is achieved by
  orientation.
• Blazars are AGNs with jets close to the line of
  sight. Includes
• BL LAC objects (which present a complete or
  nearly complete absence of lines)
• Higly polarized quasars, FSRQs (although these
  are in general more distant, luminous and have
  stronger emission lines).

• Main physical components
• Black hole Jet
• Obscuring torus surrounding the accretion disk
• Broad line region (clouds rotating at high
  velocity yield to broad emission lines)
• Narrower emission lines are produced farther from
  the BH
• A hot electron corona populates the inner region,
  probably generating X-ray emission

7
Active Galactic Nuclei Basic phenomenology
Radio to g-ray energy distribution of 3C 279 in
low and high state measured in January and
February, 1996. Wehrle et al. (1998). General
features are a) strong flux variability, b)
spectral variability, especially when flaring,
and c) the dominance of the gamma-ray emission
over all other wavelengths.
8
Flares so fast imply a beamed, small source of
gamma-rays

• Optical depth to gamma-gamma
• For a photon energy of 1 MeV, and a
  luminosity of 1048 erg s-1, the optical depth is
  t gt 200 / (tv/1 day)
• Elliot Shapiro relation for a spherical
  accretion the source luminosity is limited by
  Eddingtons and the size of the source has to be
  larger than the Schwarzschild radius
• (Indication Distance is not a problem)

9
Flares so fast imply a beamed, small source of
gamma-rays

• If the emission is beamed -gt special relativistic
  effects

10
Active Galactic Nuclei Theories with leptonic
dominance

• Radio to UV -gt Synchrotron radiation of
• relativistic electrons
• MeV-GeV component-gt Inverse Compton
• scattering of low energy photons

• Possible photons targets
• Synchrotron photons produced in the jet SSC
• UV-Soft and X-ray continuum from the disk ECD
• UV-Soft X-ray continuum after reprocessing at the
  BLR ECC
• Synchrotron radiation reflected at the BLR RS

11
Credits

• SSC or Self-Synchrotron Compton process e.g.
  Marscher Gear 1985, Maraschi et al. 1992, Bloom
  et al. 1996
• ECD or External Comptonization of Direct disk
  radiation process e.g. Dermer et al. 1992,
  Dermer Schlickeiser 1993
• ECC or External Comptonization of radiation from
  Clouds e.g. Sikora et al. 1994, Dermer et al.
  1997, Blandford and Levinson 1995
• RS or Reflected Synchrotron mechanism e.g.
  Ghisellini Madau 1996, Bottcher Bednarek
  1998, Bednarek 1998

Not exhaustive
12
In action
The low-frequency radio emission is expected to
be produced by less compact regions. Most FSRQs
are successfully modelled with dominant EC
models.
13
In action
BL Lac Mrk421
Most BL Lacs are successfully modelled with pure
or dominant SSC models.
BL LACs -gt FSRQs
Ghisellini, Fossati, Celloti, et al.
Increasing importance of the external radiation
field
14
Active Galactic Nuclei Theories with hadronic
dominance

• observed g-ray emission is initiated by
  accelerated protons interacting with ambient gas
  or lower frequency radiation.
• In PIC models developments of pair cascades in
  the jet.
• Efficiency increase with proton energy

15
Theories with hadronic dominance Collisions

• g-rays from pp from the collision of jets with
  gas clouds
• Due to the enhanced density in the BLR clouds, pp
  interactions can dominate the pg process
• in the case of PIC models where photopion
  interactions dominates the initiation of the
  cascade
• Another possible target for the jet could be the
  wind of an OB star moving through the jet.
• Protons responsible only for the injection of
  electrons, which in turn produce the observed
  g-ray emission by SSC mechanism (Kazanas
  Mastiachidis 1999). Large proton densities.

16
Credits

• PIC or proton induced cascade model e.g.,
  Mannheim Biermann 1992, Mannheim 1993 1996
• Sync. Radiation of protons and modelling of TeV
  blazars e.g. Aharonian 2000, Mucke Protheroe
  2000, Protheroe Mucke 2000
• Collisional models with gas e.g. Beall
  Bednarek 1999, Purmohammad Samimi 2001
• Collisional models with star winds e.g. Bednarek
  Protheroe 1997

Not exhaustive
17
Theories with hadronic dominanceUltra-High
Energy Cosmic Rays and Neutrinos
e.g. Neronov Semikoz 2003
Torres Anchordoqui 2004
IceCube has sensitive to detect the concomitant
neutrino signal
Auger will test the existence of the GZK cutoff
and search for correlations.
18
Simultaneous observations can also tell the
difference e.g. W Comae (LBL)
19
Active Galactic Nuclei. The multiwavelength
approach
e.g. Hartman et al. 1999 Mattox et al.
1997,2001 Bloom et al. 1997 Wallace et al.
2002, Mirabal et al. 2000, Halpern et al. 2003,
Sowards-Emmerd 2003
3EG Reminder A-AGNs were found within the 95
confidence location contours, and present a 5 GHz
radio flux around 1 Jy together with a radio
spectral index generally above -0.5. a-AGNs
are near but outside the 95 confidence contour
and the candidate counterpart has a lower radio
flux.
20
Active Galactic Nuclei. The multiwavelength
approach
e.g. Hartman et al. 1999 Mattox et al.
1997,2001 Bloom et al. 1997 Wallace et al.
2002, Mirabal et al. 2000, Halpern et al. 2003,
Sowards-Emmerd 2003

• Radio Catalogs
• 3EG catalog AGNs were largely selected from the
  Green Bank 4.85 and 1.4 GHz single dish surveys
  (Condon et al. 1991, White Becker 1992) and the
  4.85 GHz Parkes-MIT-NRAO (PMN, Griffith and
  Wright 1993) for the southern sky.
• These frequencies are used because of the
  existence of Catalogs, not based on any
  particular physical motivation.

• All EGRET blazars with peak gamma-ray flux above
  10-6 photons cm-2 s-1 were bright (S5 gt 1 Jy)
  radio sources.
• The radio/gamma-ray correlation could be
  non-linear, presenting a trend towards low
  S5/F(gt100) MeV with increasing gamma-ray flux
  (e.g. Wallace et al. 2002, Mirabal et al. 2000,
  Halpern et al. 2003, Sowards-Emmerd et al. 2003,
  Zhang et al. 2001, Cheng et al. 2000)
• 3EG J07435447, with S5272 mJy, and 3EG
  J2006-2321/PMN J2005-2310, with S5260 mJy are
  representative of a small group of EGRET blazars
  that are dim and flat at 5 GHz but have been
  found to have brighter and flatter spectra
  extending beyond 200 GHz (e.g. Bloom et al. 1997,
  Wallace et al. 2002).

Not all good EGRET blazar candidates are radio
loud.
21
Active Galactic Nuclei. The multiwavelength
approach
Mattox et al. 2001, Sowards-Emmers et al. 2003

• Mattox et al. (2001) searched (statistical
  method) for potential radio counterparts to all
  sources listed in the 3EG, allowing for sources
  (up to the extent of their catalogs) with
  arbitrarily low S5.
• 46 blazar (all A but one) identifications with a
  high probability of being correct,
• and 37 additional plausible radio associations
  (including 15 unidentified sources, none at high
  latitudes).
• No unidentified gamma-ray source at high latitude
  at least plausibly associated with a blazar under
  this identification method
• Sowards-Emmerd et al. (2003) used the 3.5 cm
  CLASS survey and a new figure of merit to
  evaluate the plausibility of the associations. 66
  out of 116 northern EGRET sources have at least
  one plausible blazar like candidate (50 more
  high confidence classifications than Mattox et
  al.)

There are a few EGRET sources at high latitude
that have more Than 1 plausible AGN counterpart,
and some of these are along the major axis of the
contour.
22
Active Galactic Nuclei. The multiwavelength
approach
Von Montigny (1995) Jorstad et al. (2001)
Marscher et al. 2002

• From the relative timing of superluminal
  ejections and gamma-ray flares it is concluded
  that these events are correlated (99.99 CL)
• The population of bright blazars detected by
  EGRET can therefore be categorized as highly
  superluminal,
• Apparent speeds are as high as 40 c (the peak of
  the distribution being at 10h-1c, higher than
  the average speed of jet components in the
  general population of strong compact radio
  sources)

23
Active Galactic Nuclei. The multiwavelength
approach
Tornikokski et al. 2002 Romero et al. 2002 Dai
2001

• Tornikokski et al. (2002) AGNs in the 3EG
  catalog are objects that are bright and variable
  in the mm domain, having a flat spectrum up to
  100 GHz.

• Romero et al 2002 Optical microvariability
  was searched for a sample of 20 southern EGRET
  AGNs, and timescales of variation of the order of
  several hours were found.

24
Active Galactic Nuclei. Microlensing in the
gamma-ray regime
Torres et al, 2000 2003

• The region where gamma-rays of energy E are
  absorbed in local radiation fields through pair
  production gamma-ray x-ray ? e grows with the
  gamma-ray energy
• As soon as the sizes of these emission regions is
  comparable to the Einstein radius of the
  interposed lenses, gravitational lensing cannot
  proceed.
• Differential amplification.

Ray tracing Light rays are tracked back towards
the source. -deflected in the lens plane
according to the lens equation. -The
magnification map number of rays per pixel is
formed. -A source with a given size moves
through this map and generates lightcurves.
25
Active Galactic Nuclei. Microlensing in the
gamma-ray regime
Torres et al, 2000 2003
Example of magnification map for an interposed
galaxy.
Each numbered trajectory generates a different
lightcurve. Trajectories 1-6 are randomly
selected. Trajectory 7 is selected on purpose
to maximize the resultant magnification. The
sizes of the different gamma-spheres for
different energies (100 MeV, 1 GeV, 100 GeV) are
depicted in the bottom-left corner.
26
Active Galactic Nuclei. Microlensing in the
gamma-ray regime
Torres et al, 2000 2003
Red 100 MeV -- Green 1 GeV -- Blue 10 GeV
3 yrs
5days
27
At high latitudes Are AGNs all there is?
Punsly 1997
Points represent the number of gamma-ray
detections for which the counterparts are beyond
the 95 confidence contour. The dotted curves
are the boundaries of the 68 confidence band for
the hypothesis that the radio sources are
randomly distributed in the EGRET detection
fields. The number of sources whose possible
counterpart are beyond the 95 confidence contour
is compatible with the chance expectation.
An statistical indication that the answer is no.
28
Are AGNs all there is?
Sowards-Emmers' et al. 2003
Product of overdensities at different frequencies
and EGRET likelihood
The significance of the Sowards-Emmers' et al.
(2003) FoM statistic Random (histogram) and
true (points, with Poisson error bars)
distributions of the FoM. The fractional excess
(true ID fraction) for each bin is shown by the
(Poisson) error range of the shaded region (right
scale). It also selects sources which are
non-blazars.
The BAD
The UGLY
The GOOD
29
Are AGNs all there is?
Torres et al. 2001, Nolan et al 2003
Indication of a more variable population of
sources at high latitude, but variability is not
strictly connected with sky position for the
unidentified detections
30
Are AGNs all there is?
Which is the percentage of the diffuse gamma-ray
background that is produced by unresolved AGNs?
Igor Moskalenkos talk
Good agreement in the community
31
Sources of radiation different from AGNs Galaxy
Clusters
Have they appeared already in gamma-ray catalogs?
Will they appear in forthcoming ones? How many?
Which ones?

• Olaf Reimers talk

Excellent agreement in the community
32
Radiogalaxies (1000 times more abundant than
blazars)
EGRET report Sreekumar et al. 1999
2MASS

• But only Cen A (FR I, b 20 D 3.5 Mpc Flux
  is 10-7 photons cm-2 s-1 , what implies a
  luminosity of 1041 ergs s-1, 105 times less
  than blazars).
• The high-energy flux of Cen A appears to be
  constant. The EGRET photon spectrum is a single
  power law (index 2.40).
• Co-spatial detections by OSSE (Kinzer et al.
  1995) and COMPTEL (Steinle et al. 1998) provide
  consistency.

• More distant members of its class would be beyond
  the reach of EGRET.

VLA
33
RadiogalaxiesTwo particular (unconfirmed)
candidates
Mukherjee et al. 2002, Combi et al. 2003
Mukherjee et al. 3EG J16218203/NGC6251 Combi et
al. 3EG J1735-1500/J1735-15
Luminosity lt 1042-44 erg/s Distance lt 100 Mpc
34
Radiogalaxies Stacking
Cillis et al. 2004

• No result above 2s for any class or subclass, for
  any number of members ordered by redshift, 5 GHz
  flux, or optical magnitude.
• Fornax A, perhaps detected, with (TS)1/22.2
• Simulations show that gamma-ray fluxes can not
  scale simply with the ratio of radio fluxes
  between the galaxy and Cen A

35
Galaxies
Fichtel 1991, Sreekumar et al. 1992, 1993

• The only galaxy, other than the Milky Way,
  detected by EGRET is the LMC (Sreekumar et al.
  1992). Its emission is interpreted as secondaries
  of cosmic ray interactions (Fichtel 1991)
• The non-detection of the SMC proves that the
  cosmic ray distribution is not universal but
  rather galaxy-related (Sreekumar et al. 1993),
  settling an issue being around for decades!
• Both are to be observed with future instruments

36
Normal Galaxies
Pavlidou Fields 2001

• However. not many normal galaxies will be
  detected by GLAST. Only M31 and M33 appear to be
  plausible candidates.

Simple estimations of Brem.Pion decay
37
Nearby Starbursts
Akyuz et al. 1992, Volk et al. 1996, Paglione et
al. 1996, Bloom et al. 1999
10 starbursts selected by distance
(lt10Mpc), Infrared luminosity (gt109 Lsolar) at
latitudes bgt10.
38
Luminous infrared galaxies
Sanders and Hernquist, 1996
Luminous infrared galaxies have LFIRgt1011 Lsun,
Ultra-luminous have LFIRgtL12 Lsun
39
Luminous infrared galaxies
Sanders and Hernquist, 1996
40
Almost all ULIRGs seems to be double or
interacting
Butonly one within the 100 Mpc sphere Arp
220 However... There are tens of LIRGs
(luminosities gt1011 LSUN) detectability depends
on the combined effect of distance and starburst
activity. Some lessons to learn from gamma-ray
astronomy cosmic ray distributions in regions of
high star formation relative influence of
central black holes.
review on LIRGs and ULIRGs Sanders and Mirabel,
ARAA, 1996
41
Detectability of LIRGs
Torres et al 2004

• Gamma-ray detectability is favored in starburst
  galaxies (Akyuz, Aharonian, Volk, Fichtel, etc)
• Large M, with high average gas density, and
  enhanced cosmic ray density
• Recent HCN-line survey of Gao Solomon (2004) of
  IR and CO-bright galaxies, and nearby spirals
• Allows estimate of SFR (from HCN luminosity) and
  minimum required k for detection by LAT and IACTs
  (from HCN CO intensities and distance)
• Several nearby starburst galaxies and a number of
  LIRGs and ULIRGs are plausible candidates for
  detection

42
Upper limits from EGRET on fluxes from LIRGs and
ULIRGs
Torres et al 2004

• 2s upper limits for some of the nearest Luminous
  Infrared Galaxies (gt1011 L in IR)
• Data from entire EGRET mission were used (LIGs
  are steady sources)
• Typical flux limit 5 10-8 cm-2 s-1

43
Analysis supported by detailed modelling (Arp
220)
Torres 2004
Dust
Radio
Blackbody optical contribution
Non-thermal radio contribution
Torres 2004
Torres 2004
IR-FIR and radio modelling of the different
components of Arp 220 with secondaries produced
in cosmic ray interactions confirm an observable
gamma-ray source.
44
From me far off, with others all too near High
latitude molecular clouds
Torres 2004b

• Population of nearby (100 pc), high
  Galactic-latitude (bgt25), molecular clouds.
  These  typically have an average of tens of solar
  masses, and about a degree in extension.
• MBM Catalog (Magnani et al. 1985 later updated
  by Magnani et al. 1996).
• Some of these clouds could be GLAST sources (or
  produce a flux needed to be taken into account
  for precise background modeling).

45
Typical properties very close and not too massive
(Target Mass/10 Solar Masses) x
(Distance/100 pc)-2 x CR enhancement gt 1 Some are
plausible sources for GLAST (even for solar
neighborhood CR spectrum)
46
High-Latitude Molecular Cloud An atypical case
HLCG 92-35
Yamamoto et al. 2003
The filament is found to consist of a hundred
molecular clouds, whose total mass is estimated
to be 1200 Msun. These observations discover
a massive cloud, HLCG 92-35, with a mass 330
Msun, corresponding to 1/4 of the total mass.
Yamamoto et al. 2003
47
Concluding remarks

• Lots of discovery opportunities at high latitudes
• Dramatically improved knowledge of the high
  energy emission of AGNs and how it correlates
  with emission at lower and higher frequencies.
  Neutrinos? Cosmic Rays?
• Gamma-rays from cosmic rays emission from the
  LMC, SMC, some normal galaxies, several
  starbursts and LIRGs, also from 1 ULIRG, high
  latitude molecular clouds, and surprises
• Galaxy clusters? Background?

Thank you!