Lag-luminosity relation in gamma-ray burst X-ray flares: a direct link to the prompt emission

arXiv:1004.1568v1 [astro-ph.HE] 9 Apr 2010 Mon. Not. R. Astron. Soc. 000, 1–19 (2010) Printed 19 November 2018 (MN LATEX style file v2.2) Lag-luminosity relation in gamma-ray burst X-ray flares: a direct link to the prompt emission R. Margutti1,3⋆, C. Guidorzi2, G. Chincarini1,3, M.G. Bernardini1, F. Genet4, J. Mao1, F. Pasotti1 1INAF Osservatorio Astronomico di Brera, via Bianchi 46, Merate 23807, Italy 2University of Ferrara, Physics Dept., via Saragat 1, I-44122 Ferarra, Italy 3University of Milano Bicocca, Physics Dept., P.zza della Scienza 3, Milano 20126, Italy 4Racah Institute of Physics, Hebrew University of Jerusalem, Israel Accepted 2010 Month day. Received 2010 Month day; in original form 2010 Month day ABSTRACT The temporal and spectral analysis of 9 bright X-ray flares out of a sample of 113 flares observed by Swift reveals that the flare phenomenology is strictly analogous to the prompt γ-ray emission: high energy flare profiles rise faster, decay faster and peak before the low energy emission. However, flares and prompt pulses differ in one crucial aspect: flares evolve with time. As time proceeds flares become wider, with larger peak lag, lower luminosities and softer emission. The flare spectral peak energy Ep,i evolves to lower values following an exponential decay which tracks the decay of the flare flux. The two flares with best statistics show higher than expected isotropic energy Eiso and peak luminosity Lp,iso when compared to the Ep,i −Eiso and Ep,i −Liso prompt correlations. Ep,i is found to correlate with Liso within single flares, giving rise to a time resolved Ep,i(t) −Liso(t). Like prompt pulses, flares define a lag-luminosity rela- tion: L0.3−10 keV p,iso ∝t−0.95±0.23 lag . The lag-luminosity is proven to be a fundamental law extending ∼5 decades in time and ∼5 in energy. Moreover, this is direct evidence that GRB X-ray flares and prompt gamma-ray pulses are produced by the same mechanism. Finally we establish a flare- afterglow morphology connection: flares are preferentially detected superimposed to one-break or canonical X-ray afterglows. Key words: gamma-ray: bursts – radiation mechanism: non-thermal –X-rays 1 INTRODUCTION The high temporal variability was one of the first properties to be attributed to the Gamma-ray burst (GRB) prompt emission in the γ-ray energy band (Klebsadel, Strong & Olson 1973). The advent of Swift (Gehrels et al. 2004) revealed that a highly variable emission characterises also the early time X-ray afterglows in the form of erratic flares. This established the temporal variability as one of the key features in interpreting the GRB phenomena. GRB 050502B and the X-ray flash 050406 (Falcone et al. 2006; Romano et al. 2006b; Burrows et al. 2005b) provided the first examples of dramatic flaring activity superimposed to a smooth decay: in particular, GRB 050502B demonstrated that flares can be considerably energetic, with a 0.3-10 keV energy release comparable to the observed prompt fluence in the 15-150 keV band. ⋆E-mail: raffaella.margutti@brera.inaf.it (RM) Thanks to the rapid re-pointing Swift capability, it was later shown that flares are a common feature of the early X-ray afterglows, being present in the ∼33% of X-ray light-curves (Chincarini et al. 2007, hereafter C07; Falcone et al. 2007, hereafter F07). On the contrary, a convincing optical flare, counterpart to a detected X-ray flare is still lacking, suggesting that the detected optical afterglow contemporaneous to the high-energy flares is dominated by a different emission component (see e.g. GRB 060904B, Klotz et al. 2008 but see also Greiner et al. 2009 where an optical flare was probably detected but, unfortunately, contemporaneous X-ray coverage is lacking). Based on the temporal and spectral study of a statistical sample of X-ray flares within GRBs, C07 and F07 showed that the flares share common properties and that the flare phenomenology can be described using averaged properties (see C07 and F07 and references therein): • The same GRB can show multiple flares (see e.g. 2 R. Margutti et al. GRB 051117A which contains a minimum of 11 structures in the first 1 ks of observation); • The underlying continuum is consistent with having the same slope before and after the flare, suggesting that flares constitute a separate component in addition to the observed continuum; • Each flare determines a flux enhancement evaluated at the peak time ∆F/F between ∼1 and ∼1000, with a fluence that competes in some cases (e.g. GRB 050502B) with the prompt γ-ray fluence. The average flare fluence is ∼10% the 15-150 keV prompt fluence; • Flares are sharp structures, with ∆t/t ∼0.1, a fast rise and a slower decay; • Each flare determines a hardening during the rise time and a softening during the decay time (F07), reminiscent of the prompt emission (e.g. Ford et al. 1995): the result is a hardness ratio curve that mimics the flare profile (see e.g. GRB 051117A, Goad et al. 2007, their figure 9). In this sense flares are spectrally harder than the underlying

…(Full text truncated)…

This content is AI-processed based on ArXiv data.