1. Properties of X- Ray Rich Gamma- Ray Bursts and X -Ray Flashes Valeria D’Alessio & Luigi Piro INAF: section of Rome, Italy XXXXth Moriond conference, Very High Energy Phenomena in the Universe March 12th-19th, 2005, la Thuile, Aosta, Italy

2. Outline 1. Introduction to X-Ray Rich Gamma-ray Bursts (XRRs) and X-Ray Flashes (XRFs) 2. Description of our analysis of the following XRR/XRF properties: - Distribution of spectral parameters of the prompt emission - Distribution of X and optical fluxes of the afterglow 3. Discussion of the results in the framework of high redshift scenario and inhomogeneous jet model off axis scenario 4. Conclusions

3. What are X Ray Flashes and X Ray Rich Gamma-Ray Bursts? X Ray Flashes (XRFs) are a subclass of GRBs, ~ 1/3 of them, discovered by BeppoSax in 2001 (Heise & in' t Zand, 2001) (fig.1) as: Events no detected by GRBM (40-700 keV) Events with high non thermic emission in X range 2-10 keV X Ray Rich GRBs (XRRs) are a subclass of GRBs characterized by: Very faint Gamma to X fluence in comparison with that of GRBs. Lamb et al. (2003) defined a criterious of classification for different events, according the value of spectral hardness ratio H=S(2,30)/S(30,400) : – GRBs: events with log H < -0.5 – XRRs: events with -0.5 < log H < 0 – XRFs: events with log H > 0 Fig1 : Light curves of GRB980329 (left) and XRF971019 (right) in range 2-28 keV and 40-700 keV (Heise 2003) Fig 2: Comparison of spectrum for a classical XRF, XRR, GRB (Lamb et al. 2003) XRF and XRR spectrum, as GRBs, is described by Band law. Kippen et al. (2001) and Sakamoto et al. (2004) found: –spectral slope α and β are marginally consistent with those of GRB –E p is significantly lower than GRB one

4. Origin of XRFs GRBs at high redshift (z >5) (Heise 2003): E p would be shifted by a factor (1+z) GRBs seen off-axis : -uniform jet model ε~ cons off-beam (Yamazaki, Ioka & Nakamura 2003) -universal structured jet model ε~ θ -2 (Zhang & Meszaros 2002, Rossi et al. 2002) - quasi-universal gaussian-like structured jet model ε~ e -θ²/2θ² o (Zhang et al. 2004, Lloyd-Ronning et al. 2003) Unified jet model (Lamb et al.2003): XRFs have a wide opening angle jet, while classical GRBs have an high collimated jet. Dirty fireball (Dermer et al. 1999): in the external Shock model, fireball with high baryon load would lead to a smaller Lorentz factor and consequently smaller E p Clean fireball (Mochkovitch et al. 2003): in the internal Shock model, fireball with the bulk Lorentz factor >> 300 and the contrast between the bulk Lorentz factors of colliding shells small could produce smaller E p Off axis cannonballs (Dar & De Rujula 2003) Photosphere-domianted fireballs (Dermer et al.1999; Huang, Dai & Lu 2002) Peripheral emission from collapsar jets (Zhang, Woosley & Heger 2003)

5. ➔ We compiled a sample of all the events observed until 31 December 2003 and classified in litterature as XRRs/XRFs. ➔ They are 54 events, 17 observed by BeppoSax and 37 observed by HETE-2. ➔ We classified them according Lamb et al. (2003) criterious established for events observed by HETE-2. We confirm that the 37 events observed by HETE-2 have log H > -0.5 We find that also the 17 events observed by Beppo Sax have log H > -0.5 The analyzed sample of XRRs and XRFs All the 54 events have log H > -0.5 and so they are XRRs/XRFs; in particular we find 26 XRFs and 28 XRRs, but we consider them as an unique class.

6. Distributions of α, β and E p We built up the distributions of spectral parameters α, β and E p of Band law for the bursts of sample which have these parameters well constrained and we compared results of their mean value with that of 31 classical GRBs, 21 observed by BeppoSax and 10 by HETE-2 (tab 1.,fig 3, 4) Fig. 3 Distribution of spectral slope   (left) and  (right) for XRRs/XRFs (red line) and GRBs (black line). Tab1.Mean value of α, β and E p for the class of XRRs/XRFs and GRBs. XRRs/XRFs GRBs CLASS <  <  XRRs+XRFs-(1.20±0.05) [37]-(2.83±0.22) [19]47±5 [51] GRBs-(0.98±0.07) [31]-(2.86±0.23) [25]194±28 [30]

7. F ig. 4: Distribution of Peak energy (right) for XRRs/XRFs(red line) and GRBs (black line). We find that : XRRs/XRFs is compatible with that of GRBs within 3σ XRRs/XRFs is compatible with that of GRBs within 1σ XRRs/XRFs is lower by a factor ~ 4 compared to that of GRBs We confirm results of Kippen and Sakamoto XRRs/XRFs GRBs

8. What we expect for X and Optical afterglow properties I n the high redshift scenario: 1. no detection of Optical afterglow for XRRs/XRFs 2. X-Ray flux of the afterglow of XRRs/XRFs fainter than GRB one. In fact, the observed flux depends on redshift as (Lamb & Reichart 2000): F(ν,t)= L ν (ν,t) / 4πD²(z)(1+z) 1-α+δ where α is the spectral index, δ is the temporal decaying index and D(z) is the comoving distance In particular: if XRR/XRF are at z =5 the ratio between their X and GRB at z=1 afterglow is ~ 7

9. In the off-axis scenario: The afterglow of XRRs/XRFs is fainter than GRB one, more and more as observing angle increases, but only at early time from burst trigger (fig. 5) In particular: in the universal stractured jet model E iso = 4  ε ∝ θ - ² with E p ∝ E iso 1/2 → E p ∝ θ -1 E p ( GRB )/E p ( XRR/XRF )=θ XRR/XRF /θ GRB From our results of E p : θ XRR/XR /θ GRB =4.13±0.67 Assuming: XRRs/XRFs and GRB at z=1 Afterglow observed 11.1 hr from burst trigger The ratio between observed afterglow flux of GRBs and XRRs/XRFs, considering time dilatation, is ~6 if θ XRR-XRF =4° and θ GRB =1° ~24 if θ XRR-XRF =8° and θ GRB =2° ~74 if θ XRR-XRF =16° and θ GRB =4° Fig 5.light curves of an inhomogeneous jet obsereved from different angle. From the top: θ  =0.5°, 1°, 2°, 4°, 8°, 16° (Rossi et al. 2002).

10. What we find for X and optical afterglow ➔ Ro = f o GRB / f o XRR/XRF =1.48±0.55 ➔ Rx = f x GRB / f x XRR/XRF =1.20±0.64 The fluxes of the X and optical afterglow of the XRRs/XRFs are consistent with that of GRBs!! Neither the inhomogeneous jet off axis scenario nor the high redshift scenario are consistent with the properties of the total XRR/XRF sample The results of mean value for the logarithm of the optical and X fluxes of XRRs/XRFs with observations within 1.5 d, compared with 27 GRBs ( De Pasquale et al. 2004) are: CLASSLog(fo) σ²Log(fx) σ² XRRs+XRFs-(31.02±0.11) 0.38-(12.32±0.21) 0.35 GRBs-(30.85±0.12) 0.39-(12.24 ±0.12) 0.26

11. Fig 9 Distribution of logarithm of X flux at 11.1 hr from burst trigger in unit erg cm -2 s -1 for 27 GRBs (black line) and 15 XRFs/XRFs (red line); HG=XRRs/XRFs with host galaxy, OT=XRRs/XRFs with optical transient, X= XRRs/XRFs with neither HG nor OT. Fig 8 Distribution of logarithm of Optical flux at 11.1 hr from burst trigger in unit  Jansky  for 11 GRBs (black line) and 10 XRFs/XRFs (red line) with optical afterglow with early observations

12. What we find for optical afterglow observations and redshift For the 54 events there are 40 events with Optical Afterglow observation: 24/40 events have no detected candidate optical afterglow 9/54 events have an estime of spectroscopic reshift; 3/9 from host galaxy and 6/9 from optical afterglow. The mean value is =1.41±0.39 3/54 events have redshift constraints: redshift is always z< 3.5 The high redshift scenario is not consistent with the global class of XRRs/XRFs

13. Conclusions We confirm spectral properties of XRFs and XRRs are similiar to those of GRBs, except the lower value of Peak Energy which is lower. We find that the X and Optical fluxes of the afterglow of XRRs/XRFs are compatible with that of GRBs and that the the mean value of redshift for 9 XRRs/XRFs is a low value. Nor high redshift scenario neither inhomogeneous jet model scenario can explain the properties of all the XRR/XRF class.

14. Analysis 1. We analysed in particular X and Optical Flux of the afterglow at 11.1 hr from burst trigger and we compared distribution and mean value of logF o and logF x for XRRs/XRFs and GRBs. 2. When the value of F o and F x at 11.1 hr was not available in literature, we extracted them at 11.1 hr from the observations of the afterglow at different time with the best temporal slope between prompt and afterglow observations. We used only events with observations until 1.5 day from burst trigger. 3. We find: 1.15/54 X candidate afterglow, 9 of them with observations at early time (4 XRF and 5 XRR) 2.40/54 optical observations with:16 Optical candidate afterglow and 24 “DARK” events, but 9 of them have not early time observations  31 events, 21 DARK (11 XRR and 10 XRF) and 10 OT (7 XRR and 3 XRF)

15. Criterious of definition Lamb et al. (2003) defined a criterious of classification for different events, according the value of spectral hardness ratio H= S(2,30)/S(30,400) : GRBs: events with log H < -0.5 XRRs: events with -0.5 < log H < 0 XRFs: events with log H > 0 Histogram of hardness ratio for GRBs ( blue), XRRs (green) and XRFs (red) observed by HETE-2 (Lamb et al. 2003)

16. Spectral properties of prompt emission of XRRs and XRFs XRFs and XRRs, as GRBs, have a spectrum described by Band law (fig. 3): E  exp(-E/E o ) E ≤ ( α -β)E o N(E) ~ E  E ≥ ( α -β)E o With E p =(2+ α) E o Kippen et al. (2001) analysed 9 XRFs observed by BeppoSaX and by BATSE off line data. They found: 1. spectal slope α and β are marginally consistent with those of GRBs 2. E p is significantly lower than GRB one, which is ~300 keV. Sakamoto et al. (2004) analysed a sample of 16 XRFs and 19 XRRs observed by HETE-2. He confirmed the results of Kippen et al. (2001). Fig 3: Comparison of spectrum for a classical XRF, XRR, GRB (Lamb et al. 2003)

17. Implication on inhomogeneous jet model (II):Isotropic Energy Distribution We calculated isotropic energy of the 14 XRRs/XRFs with estimated redshift and we compared them with values of 17 GRBs (Bloom et al, 2001). We obtained mean value for this parameter and distribution (fig.9): ● = (46±18)10 51 erg for XRRs/XRFs ● = (330±100) 10 51 erg for GRBs This results is consistent within 1 , BUT there are three events, XRR000615, XRR011030 and XRF020903, whose E iso is lower by 3 or 4 order of magnitude compared to GRB one. Fig. 9: Distribution of logarithm of isotropic Energy for XRRs/XRFs (red line) andRBs (black line). G E iso (GRB)/ E iso (XRR/XRF) =7.17 ± 3.55 Since E iso = 4     - ² R iso = Eiso GRB / Eiso XRR/XRF = (  GRB /  XRR/XRF ) -2 With previous assumption and results we obtain R iso =4±8

18. What are X Ray Flashes and X Ray Rich Gamma-Ray Bursts? Events no detected by GRBM (40-700 keV) Events with high non thermic emission in X range 2-10 keV X Ray Flashes (XRFs) are a subclass of GRBs, ~ 1/3 of them, discovered by BeppoSax in 2001 (Heise & in' t Zand, 2001), as: X Ray Rich GRBs (XRRs) are observed as: Events detected in Gamma range by GRBM Events with very faint Gamma to X fluence Fig1: Light curves of GRB980329 (left) and XRF971019 (right) in range 2-28 keV and 40-700 keV (Heise 2003)

19. ➔ Appling this result of Rx = f x GRB / f x XRR/XRF we extracted the value of the observing angle of XRRs/XRFs, θ XRR/XRF, assuming θ GRB =1°.  θ XRR/XRF = (2 -2 +2 )° Implication on inhomogeneous jet model : observing angle The inhomogeneous jet scenario is not consistent with global class of XRRs/XRFs