F.Longo SNR 1 GLAST LAT GLAST Italian SW meeting 11 march 2004 SNR Francesco Longo University and INFN, Trieste, Italy [email protected] Slides.

F.Longo SNR 1

GLAST LAT GLAST Italian SW meeting 11 march 2004

SNRSNR Francesco LongoUniversity and INFN, Trieste, [email protected]

Slides from Diego Torres (Livermore)In collaboration also with M.Ajello (Monaco) and R.Rando (Padova)

mailto:[email protected]

F.Longo SNR 2


Raggi Cosmici Galattici e SNRRaggi Cosmici Galattici e SNR

F.Longo SNR 3


SNR al GeVSNR al GeV

• SNR rivelabili tramite IC, Bremsstrahlung e decadimento del π0

• Flusso atteso -> ~ 10-7 ph cm-2 s-1

• Fondo diffuso -> ~ 10-7 ph cm-2 s-1

SNRdifficili da rivelare

Shock front

Molecular cloud

protons, electrons

Synchrotron radiation

e

p

F.Longo SNR 4


Origine dei CR in astronomia gammaOrigine dei CR in astronomia gamma

• Trovare traccia d’interazione adronica

nello spettro di una sorgente• Nessuna sorgente del III Catalogo EGRET mostra tale

caratteristica• Esiste prova della propagazione dei CR-adronici, ma non della

loro accelerazione

F.Longo SNR 5


Spettro da decadimento del Spettro da decadimento del ПП00

Propagazione dei raggi cosmici

Spettro da decadimento di 0

Spettro del piano galattico

F.Longo SNR 6


Possibilità di rivelare SNR al GeVPossibilità di rivelare SNR al GeV• Canale adronico incrementato !

• FSNR ~ ρISM Fnubi ~ εCR = 1eV/cm3

• Nubi vicino SNR εCR>> 1eV/cm3

canale adronico incrementato• R= CO(J=2->1)/CO(J=1->0)• R~0.7, ma R2.5 per nubi eccitate

SNR

Nube

F.Longo SNR 7


Osservazione multi-bandaOsservazione multi-banda

• SNR RX J1713.7-3946

Sorgente GeV rivelata da

EGRET

Nube in interazione col residuo

Sorgente TeV rivelata da CANGAROO

Residuo di supernova rivelato in X da ASCA

F.Longo SNR 8


Where to produce gamma in the galaxy?Where to produce gamma in the galaxy?

Shock front

Molecular cloud

protons, electrons

Synchrotron radiation

e

p

Nearby molecular clouds can provide targets for ions accelerated at the SNR shock. Gamma-rays are then produced by neutral pion

decay pointing out the production of hadronic cosmic rays (e.g. Aharonian et al. 1994: A&A 285, 645).

F.Longo SNR 9


CRs below the “knee” are thought to be accelerated at supernova remnats (Syrovatskii 1953). There, charged particles would be accelerated by Fermi mechanism operating at the strong blast wave shock (Bell 1978).

Synchrotron emission detected from radio to X-ray energies in SNRs clearly shows the presence of TeV leptons in these sources. Direct emission from locally accelerated hadrons, on the contrary, cannot be observed.

Since CRs are deflected by the galactic magnetic field, they do not preserve the information on the location of their source. We must, consequently, look for electromagnetic signatures produced by the protons and ions during their accleration.

Gamma SNR?Gamma SNR?

F.Longo SNR 10


A case by case analysisA case by case analysis

• A case by case analysis is needed for each SNR-EGRET source coincident pair.

• There should be, nearby, enhancements of molecular material that could act as target for accelerated protons.

• This material, then, must be excited by the shock.• Leptonic processes and other candidate sources must be

discarded as the origin of the gamma-ray radiation.

Torres et al. astro-ph/0209565, Supernova Remnants and gamma-ray sources, Review for the Physics Reports (2002)

F.Longo SNR 11


The most interesting case? The most interesting case? [G347.3-0.5][G347.3-0.5]

• Positional coincidence of the non-variable EGRET gamma-ray source, 3EG J1714-3857, with a very massive (~3×105 solar masses) and dense (~500 nucleons cm-3) molecular cloud…

• This molecular cloud is interacting with the X-ray and TeV gamma-ray emitting SNR G347.3-0.5…

• The cloud region is near the shell of the SNR, and shines at GeV, but it is of low radio and X-ray brightness…

Initial discussion: Butt, Torres, et al. ApJ Letters, 562, 167 (2001)Recent results: Butt, Torres, et al., Nature 418, 499 (2002)

Slane et al. ApJ 525, 357 (1999)

The clouds that seem to interact with it are pushed away as a cause of the blast wave shock

Total molecular column density over a wide section of the fourth Galactic quadrant around G347.3-0.5. The lowest contour is well above the instrumental noise (9) to emphasize the relatively low molecular column density toward the SNR.

Molecular environment of the SNR G347.3-0.5

F.Longo SNR 13


The distribution of 781 line intensity ratios, R={CO(J=21)/CO(J=10)}, measured every 15′ in the region from l=346.5348.5; b= -0.5+0.5, and averaged over 5km/sec bins of velocity between vlsr= -150 km/sec +50

km/sec.

The mean of the distribution, ~0.72, agrees with the average unexcited value in the Galactic plane. The cloud however, show values 3 above that value.

More precise indication of interaction with molecular material

Top 0.5% of all values measured. All other bins with high R are well outside the 3EG field

F.Longo SNR 14


The complete panorama for SNR G347.3-0.5

ROSAT X-ray contours. Emission from the bulk of the SNR rim can be seen with particular enhancements along the west/northwest regions, where bright non-thermal radio emission is also seen. The total radio flux is well below 10 Jy, Slane et al. ApJ 525, 357 (1999)

Red depicts the TeV significance contours. The flux was (5.3 ± 0.9 [statistical] ± 1.6 [systematic]) x 10-12 photons cm-2 s-1 (at E>1.8 ± 0.9 TeV). Muraishi et al. AA354, L57 (2000).

While electrons give rise to the bulk of the non-thermal radio, X-ray and TeV emission in the NW, the CR protons and ions are exposed at GeV energies via their hadronic interactions in the dense material of

cloud A, leading to pion gamma-decay GeV emission in the NE.

F.Longo SNR 15


The expected -ray flux at Earth coming from the SNR is (Drury et al. 1994),

ESN is the energy of the SN in ergs, is the fraction of the total energy of the explosion converted into CR energy, and n and d are the number density and distance. In most cases, this flux is far too low to be detected by EGRET, but the existence of massive clouds in the neighborhood can enhance the emission

Here M is the mass of the cloud in thousands of solar masses, k is the CR enhancement out of the usual emissivity (~2.2 10-25 s-

1 H-atom-1).

0

23

910)100( kqDMMeVEF kpc

The gamma-ray luminosity

We have calculated the expected gamma-ray luminosity using the following information:

explosion energy = (1.7-2.2)×1051 ergs distance to the SNR = 6.3 ± 0.4 kpc unshocked ambient density, no= 0.01-0.3 cm-3

The mass of Cloud A, centered at (l,b)=(347.9,-0.25), is ~3×105 Mo and its mean density ~500

nucleons cm-3.

The total gamma-ray luminosity is:

Ftot(E>100MeV) = Fsnr(E>100MeV) + Fcloud A(E>100MeV)

The cosmic ray enhancement factor, ks is computed to be in the range 30-40.

Then, Ftot(E>100MeV) = (4-7) ×10-7 photons cm-2 sec-1 , with the contribution of Cloud A

dominating the GeV flux by over 2 orders of magnitude. This predicted flux is fully consistent with the measured value: (4.36 ± 0.65)×10-7 photons cm-2 s-1.

(Ellison et al. 1999)

The gamma-ray luminosity

The single power-law fit (=-2.3) through all points (solid black line) is not at ease with the enhancement at 50-70 MeV.

This feature is consistent with the long-sought SNR neutral pion gamma-decay resonance centered at 67.5 MeV.

The red curve is an expected spectrum due to hadronic CR interactions.

As Schlickeiser has pointed out, the bremsstrahlung from secondary electrons due to the decay of hadronically produced charged pions, ± ’s, will contribute significantly at energies lower than ~70 MeV.

Schlickheiser 1982

The spectrum of the EGRET sourceThe gamma-ray spectrum: consistent with hadronic production

However: deviationfrom other points is lessthan 3ust yet be

Confirmed.

The electron flux needed to explain the GeV emission via e- bremsstrahlung in the cloud material should also produce an enhanced synchrotron radio emission. The expected ratio of GeV bremsstrahlung

flux to radio synchrotron flux is:

, s cm Jy )(

103.4

)

)MeV 100( 1-2-1-2/)1(H

2/)1(μGcm

21

Jy

3

p

zpBn

pcF(ν

EFR

),( )1( )102.3( 10 22/)1(15)1(5 papc(p) pp

Measured from CO obs.

Measured from TeV obs..

Frequency of observations

Spectral index

Observed by EGRET

This is what we want: radio fluxprediction if the flux is leptonic

One thing is sure: GeV emission is not leptonic

F.Longo SNR 19


The synchrotron radio spectrum which would be expected from Cloud A, under the assumption that the GeV flux were due to electron bremsstrahlung.

Since this spectrum violates the observed upper limit (blue) by a factor of 20 at 843 MHz, we rule out a predominantly electronic origin of the GeV luminosity. (An assumed low frequency turnover at ~100 MHz is shown by the red dotted line.)

Observed upper limits

One thing is sure: GeV emission is not leptonic II

F.Longo SNR 20


No other plausible candidate in the 3EG fieldNo other plausible candidate in the 3EG field

• There are two recently discovered pulsars within the 95% confidence location contours of 3EG J1714-3857: PSR J1715-3903 and PSR J1713-3844

• Their spin down luminosity is such that they cannot contribute significantly to the observed gamma-ray emission (Torres et al. ApJ Letters, 560, 155, 2001).

• Two other SNRs within the EGRET 95% contours: CTB37A&B. They can both be ruled out as strong gamma-ray emitters because of their large distance (11.3 kpc) and the low density medium around them

• No WR or Of massive stars in the field, no X-ray binaries or black hole candidates

Torres et al. ApJ Letters, 560, 155, 2001

F.Longo SNR 21


Strong hints that the blast wave shock of SNR G347.3-0.5 is a site of hadronic cosmic ray acceleration

• TeV cosmic ray electrons are accelerated in this SNR; • the abutting cloud material is extremely excited; • the cloud region is of low radio and X-ray brightness; • the GeV flux is non-variable and in agreement with that

expected from o gamma-decays; • the spectral index is as expected for an hadronic CR

source population • there are no other candidate GeV sources within the

95% location contours of 3EGJ1714-3857

SummarySummary

Resolving the region of gamma-ray emission

F.Longo SNR 23


Emissione compositaEmissione composita

pulsar

Nube molecolare

F.Longo SNR 24


Analisi dei profiliAnalisi dei profili

Profilo in longitudine

Profilo in latitudine

F.Longo SNR 25


Analisi dei profili (IIb)Analisi dei profili (IIb)

Conteggi

Fondo diffuso

Sottrazione

F.Longo SNR 26


ConclusionsConclusions

There is a clear hint of an unambiguous connection between unidentified EGRET sources and SNRs. This will lead to the observational definite proof that TeV protons are acelerated in SNR shocks.

It is at least plausible that EGRET has detected distant (more than 6 kpc) SNRs. There are 5 coinciding pairs of 3EG sources and SNRs for which the latter apparently lie at such high values of distance and for all these cases, we have uncovered the existence of nearby, large, in some cases giant, molecular clouds that could enhance the GeV signal trough pion decay.

AGILE and GLAST, would greatly elucidate the origin for these 3EG sources, since even a factor of 2 improvement in resolution would be enough to favor or reject the SNR connection.

Torres et al. astro-ph/0209565, Supernova Remnants and gamma-ray sources, Review for the Physics Reports (2002)Butt, Torres, Romero, Dame, & Combi, Nature 418, 499 (2002)

F.Longo SNR 1 GLAST LAT GLAST Italian SW meeting 11 march 2004 SNR Francesco Longo University and INFN, Trieste, Italy [email protected] Slides.

Documents

longo snr

gamma snr

snr shock

snr g347

rivelare snr

snr difficili

snr francesco longo

nubi eccitate snr nube