What can we conclude from UHECR observations

1
What can we conclude from UHECR observations?

• Glennys R. Farrar
• Center for Cosmology and Particle Physics
• New York University

2
Key (secure) Ingredients

• Spectrum
• Composition
• Clusters of UHE events
• Correlations with Candidate Sources

gt What can be inferred alread gt Near-term
expected improvements
3
Spectrum

• Absolute energy scale still uncertain
• Air Fluorescence-Surface discrepancy 30?
• Awaiting definitive air fluorescence yields (Sept
  07?)
• Air shower simulations systematic problems
• Berezinsky et al normalizing Es by position of
  dip gives consistent spectra
• Spectrum drops off at highest energies
• No need to worry about GZK violation
• But premature to claim that GZK cutoff has been
  confirmed (absolute E scale, max. acceleration?)

4
Energy losses produce characteristic dip in
spectrum Berezinsky et al ICRC 07

• Nucleon CMB photon gt ee- production gt energy
  loss
• As long as only protons (lt 10 nuclei) spectrum
  develops a dip at a definite energy.
• Use dip to normalize energy of different
  experiments.

5
Rescaling energies to dipgt consistency!

• Left published spectra x E3 from different
  expts
• Right spectra after shifting energies by
  factors
• ?AG 0.9 ?HR 1.2 ?Yk 0.75 ?Auger
  1.2
• Shift gt consistent normalizations of all
  spectra, for ?Auger 1.4.
• (?HR 1.2 difference in AF yield
  assumptions gt ?Auger 1.35)

6
Composition

• Strong limit on photons (Auger ICRC07)
• lt 10 at 20 EeV
• Excludes non-contrived top down models
• Very confusing situation on nucleon vs. nucleus
  question!

7
Elongation Rate nowwell measured (Auger)

• Xmax increases with energy.
• For fixed total energy, Xmax is lower for nuclei.
• Predicted Xmax vs E depends on UHE event
  generator.
• Data suggests heavy ?light during
  galactic-extragalactic transition
• Then, light ? heavy at still higher energies!
• This is VERY unexpected!

8
There is a problem with present simulations!

• If the energy of a hybrid event is fixed
  according to the AF method, the predicted number
  of muons in the shower is too low by 40 other
  similar effects.
• EPOS increases the mu/e ratio but gets other
  features wrong.
• Some aspect(s) of the UHE hadron interaction are
  not yet understood, possibly
• inclusive nuclear modeling?
• Leading particle effects? (MINOS mu/mu- is
  wrong)
• ???
• May be responsible for energy normalization
  discrepancies, seemingly puzzling elongation
  rates.

9
Clusters of UHE events

• Most UHECRs are isolated on few o scale
• AGASA 96 (57 events gt 40 EeV) 1 triplet, 5
  pairs
• AGASA HiRes 04,05
• AGASA triplet promoted to quadruplet with HE
  HiRes data (37 events)
• to quintuplet w all published HiRes data (234
  events)
• Promotion probability 2 10-3
• Auger sees no statistically significant small
  scale clustering signal (ICRC07), but
• Toward galactic center
• Much more extragalactic structure

More severe magnetic deflection?
10
Extragalactic Magnetic Fieldsmay be low in some
directions, large in others Dolag et al
astro-ph/0310902, 0410419 Sigl et al
astro-ph/0302388,0401084
Deflection map of CRs above 4 1019 eV, within
100 Mpc. EGMF deflections may be small except in
isolated directions
11
Possible Acceleration Mechanisms

• Cataclysmic event, e.g., gamma ray burst (GRB),
  massive star collapse to black hole, magnetar
  birth
• Jets in AGNs (accreting supermassive BH), e.g.,
  BL Lacs, powerful radio galaxies,
• Gradual acceleration in large scale magnetic
  shocks taking millions of years
• New Physics -- e.g., decay of invisible,
  super-heavy particle created in early moments of
  Big Bang (mostly excluded now by photon limits).

12
Inferring Source(s) from bias

• Simulated Large Scale Structure(Millenium Run,
  Springel et al 2004)
• Clustering of UHECR (if from galaxies)
• (Berlind, Farrar 2007)
• Powerful because insensitive to magnetic
  deflection
• Can only discriminate characteristic mass of
  source, not specific source type
• (e.g., GRB, AGN have same bias)
• 1000 UHECRs

13
BLLac-UHECR correlationGorbunov et al,
astro-ph/0406654 Abbassi et al,
astro-ph/0507120 R. Jansson GRF, ICRC07

• 130 BLLacs in HiRes domain
• 15 correlate with 1 UHECR (7 if random)
• 2 correlate with 2 UHECRs
• 10 excess correlations, probability lt 10-4

14
Bursting vs Continuous Source???

• Spectrum discriminates!
• Bursting source
• At any given time, factor-2 spread in energies.
• But beware

15
GZK energy loss DISTORTS A CONTINUOUS SOURCE!

• Observed Spectrum of Continuous Source with GZK
  (dN/dEsrc Esrc-p)
• Green no GZK losses
• Red with GZK losses

16
For Each Cluster

• Estimate source distance from energies (use GZK)
• Spectrum of individual source
• Include GZK distortions
• Compare observed energies with prediction of
  continuous and bursting sources.
• Fit angular distribution
• Different for bursting and continuous cases
• Can infer for each cluster
• Flux (if continuous) or total energy of burst
• lt B2 x magnetic coherence lengthgt
• If bursting, delay time since burst

17
The Ursa Major UHECR Cluster 5 events from AGASA
HiRes Galaxies from SDSS R. Abassi et al
(GRF and HiRes), ApJ 623, 264 (2005) GRF
astro-ph/0501388,GRF, A. Berlind, D. Hogg
astro-ph/0507657, ApJ 642, L89(2005)
18
Probability Distribution of Source Distancefor
Ursa Major Cluster

• Figure
• Assuming mean GZK energy loss
• Solid unrenormalized energies,
• Dashed BGG renormalized energies
• Source distance 100-150 Mpc well compatible with
  matter distribution as determined from SDSS (GRF,
  Berlind, Hogg 05)

19
UHECR multiplets

• Cluster candidate in published AGASA-HiRes data
• 5 events, chance probability 2 10-3
• Energies (15), 38, 53, 55, 78 EeV (1 EeV 1018
  eV)
• (b, l) (55o,145o) (Ursa Major Cluster)
• Projecting from AGASA-HiRes suggests with a
  dataset like Auger
• 1 multiplet with 8-10 UHECRs
• ? 1 multiplet with 6 UHECRs
• 5 multiplets with 4 UHECRs
• Need good analysis tools to avoid using false
  multiplets. Maximum Likelihood method has been
  developed (GRF)
• If (when!) such multiplets are found, they will
  give a powerful constraint on GMF and sources.

20
Conclusions

• UHECRs are an astrophysical phenomenon.
• Large statistics gt many powerful tools can be
  used.
• Near-term
• Infer sources from bias, correlations, and
  cluster properties
• Medium term
• UHE/nuclear modeling, accurate event energies, gt
  statistics
• Composition, GZK
• Longer term
• Astrophysics, UHE particle physics
• UHECRS will soon become an astrophysical UHE
  particle physics TOOL!