The complex light-curve of the afterglow of GRB071010A
S. Covino, P. D'Avanzo, A. Klotz, D.A. Perley, L. Amati, S. Campana, G. Chincarini, A. Cucchiara, V. D'Elia, D. Guetta, C. Guidorzi, D.A. Kann, A. Küpcü Yoldaş, K. Misra, G. Olofsson, G. Tagliaferri, L.A. Antonelli, E. Berger, J.S. Bloom, M. Böer, C. Clemens, F. D'Alessio, M. Della Valle, S. di Serego Alighieri, A.V. Filippenko, R.J. Foley, D.B. Fox, D. Fugazza, J. Fynbo, B. Gendre, P. Goldoni, J. Greiner, D. Kocevksi, E. Maiorano, N. Masetti, E. Meurs, M. Modjaz, E. Molinari, A. Moretti, E. Palazzi, S.B. Pandey, S. Piranomonte, D. Poznanski, N. Primak, P. Romano, E. Rossi, R. Roy, J.M. Silverman, L. Stella, G. Stratta, V. Testa, S.D. Vergani, F. Vitali, F. Zerbi
MMon. Not. R. Astron. Soc. , 1–11 (2008) Printed 8 November 2018 (MN L A TEX style file v2.2)
The complex light-curve of the afterglow of GRB 071010A (cid:63)
S. Covino, † P. D’Avanzo, , A. Klotz, , D.A. Perley, L. Amati, S. Campana, G. Chincarini, , A. Cucchiara, V. D’Elia, D. Guetta, C. Guidorzi, , D.A. Kann, A. K¨upc¨u Yolda¸s, K. Misra, , G. Olofsson, G. Tagliaferri, L.A. Antonelli, E. Berger, J.S. Bloom, M. B¨oer, C. Clemens, F. D’Alessio, M. Della Valle, , , S. di Serego Alighieri, A.V. Filippenko, R.J. Foley, D.B. Fox, D. Fugazza, J. Fynbo, B. Gendre, P. Goldoni, , J. Greiner, D. Kocevksi, E. Maiorano, N. Masetti, E. Meurs, , M. Modjaz, E. Molinari, A. Moretti, E. Palazzi, S.B. Pandey, S. Piranomonte, D. Poznanski, N. Primak, P. Romano, E. Rossi, R. Roy, J.M. Silverman, L. Stella, G. Stratta, V. Testa, S.D. Vergani, , F. Vitali, F. Zerbi, INAF/Osservatorio Astronomico di Brera, via Bianchi 46, 23807, Merate (LC), Italy Universit`a dell’Insubria, Dipartimento di Fisica e Matematica, via Valleggio 11, 22100 Como, Italy Observatoire de Haute-Provence, 04870 Saint-Michel l’Observatoire, France CESR, 9 Avenue colonel Roche, Universit´e de Toulouse, 31400 Toulouse, France Astronomy Department, University of California, 445 Campbell Hall, Berkeley, CA 94720-3411, USA INAF/Istituto di Astrofisica Spaziale e Fisica Cosmica di Bologna, via Gobetti 101, 40129 Bologna, Italy Universit`a degli Studi di Milano, Bicocca, Piazza delle Scienze 3, 20126, Milano, Italy Department of Astronomy and Astrophysics, Pennsylvania State University, USA INAF/Osservatorio Astronomico di Roma, via Frascati 33, 00040 Monteporzio Catone (Roma), Italy Th¨uringer Landessternwarte Tautenburg, Sternwarte 5, 07778 Tautenburg, Germany Max-Planck-Institut f¨ur extraterrestrische Physik, Giessenbachstrasse 1, 85748 Garching, Germany Aryabhatta Research Institute of observational sciences (ARIES), Manora Peak, Nainital 263 129, India Inter University Center for Astronomy and Astrophysics, Post Bag 4, Ganeshkhind, Pune 411 007 India Stockholm Observatory, Roslagstullsbacken 21, 10691 Stockholm, Sweden Observatories of the Carnegie Institution of Washington, 813 Santa Barbara Street, Pasadena, CA 91101, USA INAF/Osservatorio Astronomico di Capodimonte, Via Moiariello 16, 80131 Napoli, Italy Icranet, International Center for Relativistic Astrophysics Network, Piazza Repubblica 10, Pescara, Italy European Southern Observatory, Garching bei Munchen -Germany INAF/Osservatorio Astrofisica di Arcetri, Largo Enrico Fermi 5, 50125 Firenze, Italy Dark Cosmology Centre, Niels Bohr Institute, University of Copenhagen, Juliane Maries vej 30, 2100 Kobenhavn, Denmark Laboratoire d’Astrophysique de Marseille/CNRS/Universit´e de Provence, 13376 Marseille Cedex 12, France Laboratoire Astroparticule et Cosmologie, 10 rue A. Domon et L. Duquet, 75205 Paris Cedex 13, France Service d’Astrophysique, DSM/DAPNIA/SAp, CEA-Saclay, 91191 Gif-sur-Yvette, France Dunsink Observatory - DIAS, 31 Fitzwilliam Street, Dublin 2, Ireland School of Physical Sciences and NCPST, Dublin City University, Dublin 9, Ireland JILA, University of Colorado, 440 UCB Boulder, CO 80309-0440, USA ASI Science Data Center (ASDC), via G. Galilei, 00044 Frascati, Italy
ABSTRACT
We present and discuss the results of an extensive observational campaign devoted toGRB 071010A, a long-duration gamma-ray burst detected by the
Swift satellite. Thisevent was followed for almost a month in the optical/near-infrared (NIR) with varioustelescopes starting from about 2 min after the high-energy event.
Swift -XRT observa-tions started only later at about 0.4 d. The light-curve evolution allows us to single outan initial rising phase with a maximum at about 7 min, possibly the afterglow onset inthe context of the standard fireball model, which is then followed by a smooth decayinterrupted by a sharp rebrightening at about 0.6 d. The rebrightening was visible inboth the optical/NIR and X-rays and can be interpreted as an episode of discreteenergy injection, although various alternatives are possible. A steepening of the af-terglow light curve is recorded at about 1 d. The entire evolution of the optical/NIRafterglow is consistent with being achromatic. This could be one of the few identi-fied GRB afterglows with an achromatic break in the X-ray through the optical/NIRbands. Polarimetry was also obtained at about 1 d, just after the rebrightening andalmost coincident with the steepening. This provided a fairly tight upper limit of 0.9%for the polarized-flux fraction.
Key words: methods: observations, gamma-rays: bursts, X-rays: individuals:GRB 071010A (cid:63)
Based on observations made also with ESO Telescopes at the La Silla and Paranal Observatory under programmes IDs 080.D-0229, 080.D-0250 and 080.D-792.c (cid:13) a r X i v : . [ a s t r o - ph ] A p r S. Covino et al. † E-mail: [email protected]
The
Swift satellite (Gehrels et al. 2004) has revolutionisedgamma-ray burst (GRB) science by introducing dedicatedobservational strategies and providing a wealth of high-quality data. In particular, a major breakthrough has beenthe spectral and temporal coverage at X-ray energies froma few minutes up to weeks or months after the GRB. Thishas provided unprecedented information on the late promptand early afterglow evolution (Nousek et al. 2006). Opti-cal observations from the ground, however, have not alwaysbeen of comparable quality and late-time phases have beenseldom monitored accurately (Covino et al. 2006). This is aconsequence of the strong increase in the number of eventsto be followed, the large average redshift of
Swift
GRBs(Berger et al. 2006; Jakobsson et al. 2006), coupled with thedifficulties in performing productive follow-up observationsfrom the ground. While the early-time afterglow is often wellsampled by the network of small and medium-class robotictelescopes around the world, the late-time afterglow can onlybe monitored with the largest instruments and only at thecost of significant observing time (Dai et al. 2007). Basedon these considerations, we started a new program of af-terglow observations with the explicit goal of providing fullcoverage of the optical/near-infrared (NIR) afterglow lightcurve for a selected number of events (building on visibil-ity and availability of early-time observations) rather thanproviding sparse datapoints for most of the observable
Swift
GRBs.The importance of dense temporal and spectral sam-pling is particularly crucial for the identification and inter-pretation of the different features in the light-curve evolu-tion that are observed both in the optical (Kann et al. 2007)and at higher energies (Zhang & M´esz´aros 2004). In partic-ular, in spite of intense observational efforts and continu-ous theoretical interest (Burrows & Racusin 2006; Covinoet al. 2006; Curran et al. 2007; Dado et al. 2007; Dai et al.2007; Panaitescu 2008, e.g.,), the search for light-curve jetbreaks fully satisfying the requirements of the standard fire-ball model, and thus allowing a correct inference of the trueenergy content of a GRB, is still far from complete (Lianget al. 2007).The GRB 071010A afterglow evolution was followed inthe optical by TAROT and in the NIR by REM and GRONDat early times, then by Gemini-N, GROND, NOT, TNG,NTT, Keck-I, Sampurnanand, and VLT in the subsequenttwo weeks. Multi-band observations were acquired to studythe temporal evolution of the afterglow spectral energy dis-tribution (SED). A polarimetric observation was carried outalmost in coincidence with a slope transition in the lightcurve.
Swift
XRT and BAT data were also studied. In Sect. 2we report information about GRB 071010A, in Sect. 3 ourobservations, reductions, and data analysis are described,and in Sect. 4 a full discussion is presented. Our main con-clusions are summarised in Sect. 5.
GRB 071010A was detected by
Swift on Oct. 10, 2007 at03:41:12 (Moretti et al. 2007) (UT dates are used through-out this paper). The BAT light curve was relatively broad, c (cid:13) , 1–11 he complex light-curve of the afterglow of GRB 071010A with a T duration of 6 ± . ± . × − erg cm − (un-certainty 90% confidence, Krimm et al. 2007). Automaticslewing to the GRB position with the Swift narrow-fieldinstruments was disabled, so the XRT observations were de-layed by 34 ks (Guidorzi et al. 2007b). The optical afterglowwas promptly identified by TAROT ( α J2000 = 19:12:14.624, δ J2000 = − z ≈ .
98. A refined analysis was car-ried out by D’Elia et al. (2008), deriving z = 0 . ± . σ error) for the absorbing system with the highest red-shift. Lines of Mg II (2796–2803 ˚A), Mg I (2852 ˚A), Si II(2336 ˚A), Fe II (2367, 2374, 2586, 2600 ˚A), and Mn II (2594,2606 ˚A) were identified. The afterglow was not detected inthe radio band at 8.46 GHz almost two days after the GRB(Chandra & Frail 2007). Late-time observations led Perleyet al. (2007) to suggest that the afterglow decay underwenta (possibly achromatic) steepening about one day after theburst. Galactic dust absorption along the GRB line of sightis E B − V ≈ .
098 mag (Schlegel et al. 1998).Throughout this paper, the time decay and energy spec-tral indices α and β are defined by F ( t, ν ) ∝ ( t − t ) − α ν − β ,where t is the trigger time of the burst. We assume a ΛCDMcosmology with Ω m = 0 .
27, Ω Λ = 0 .
73, and h = 0 .
71. Atthe redshift of the GRB the luminosity distance is 6 . ∼ . × cm), corresponding to a distance modulus µ = 44 . σ unless stated oth-erwise. The afterglow of GRB 071010A was observed by the TAROTtelescope (Bringer et al. 1999) with the white and red fil-ters, by the REM telescope (Zerbi et al. 2001) equippedwith the REMIR NIR camera (Vitali et al. 2003; Conconiet al. 2004), and by the 2.2 m MPI/ESO telescope equippedwith GROND (Greiner et al. 2008) starting from 124 s af-ter the GRB (86 s after reception of the GCN alert). Theobservations began at an initial airmass of about 2.3 withthe target already setting. In order to obtain an accept-able signal-to-noise ratio (S/N) we needed to bin the REMdata, thus decreasing the highest time resolution affordedby the instrument. Later NIR observations were carried outwith Gemini-N equipped with NIRI, the TNG equipped withNICS, and the NTT equipped with SofI. In the optical theafterglow was again observed a few hours after the GRBwith the Keck-I telescope equipped with LRIS, the Sam-purnanand telescope, the NOT equipped with PolCor, andthe VLT with FORS1 and FORS2. Polarimetry was carriedout with the VLT equipped with FORS1 on Oct. 11, 2007at 1:11:33 (0.89607 d after the GRB) with a total exposuretime of 1800 s.Data reduction was carried out following standardrecipes with the
Eclipse package (Devillard 1997). Photom- etry was computed by aperture and profile-fitting photome-try with the
Daophot (Stetson 1987) and
SExtractor (Bertin& Arnouts 1995) packages. Photometric calibration was de-rived from the observation of standard stars in the optical ontwo different nights and with the 2MASS catalogue (Strut-skie et al. 2006) in the NIR, although in a few cases thesmall field of view of NIR detectors made the calibrationmore difficult. TAROT data were calibrated with observa-tions of two field stars and final magnitudes are expressedin the R2 system of the USNO-B1 catalogue (Monet et al.2003). The results from the photometric observations arereported in Table 1.For polarimetry, the offset between instrumental andintrinsic polarization was fixed by the observation of a po-larized standard star. Instrumental polarization was derivedby observation of non-polarized stars. The procedure we fol-lowed for the analysis of the FORS1 polarization data wasextensively discussed by Covino et al. (1999, 2002, 2003).At the time of the observation the afterglow optical emis-sion was consistent with null polarization with a 95% upperlimit of 0.9% (1.3% at 3 σ ) for the degree of polarized flux.The Swift
XRT data were reduced using the xrtpipeline task (v.0.9.9), applying standard calibration and filteringcriteria, i.e., we cut out temporal intervals in which the CCDtemperature was above − ◦ C and removed hot and flick-ering pixels. An on-board event threshold of ∼ . xrtmkarf . Optical/NIR and X-raylight curves are shown in Fig. 2.The Swift
BAT data were also analyzed. The 15–150 keV mask-weighted energy spectrum was extracted withthe ftool batbinevt using the BAT refined position (Krimmet al. 2007). All the required corrections were applied: weaccounted for the slewing through the ftool batupdatephakw and produced the detector-response matrix with batdrmgen .The spectrum was corrected for systematics depending onenergy with the ftool batphasyserr and was finally groupedby imposing a 3 σ threshold on each grouped energy channel. The optical/NIR light curve is complex, with an initialrising phase, a maximum, then a decay interrupted by arather sharp rebrightening, followed by a final steepening.We choose to fit the light curve with a Beuermann function(Eq. 1, Beuermann et al. 1999) for the first part down to therebrightening: F ( t ) = A/ [( t/t b ) κα r + ( t/t b ) κα d ] /κ , (1)where A is a normalisation constant, α r(d) is the slope ofthe rise (decay) phase, and κ is a smoothness parameter.The time at which the curve reaches its maximum is t max = t b ( − α r /α d ) / [ κ ( α d − α r )] .We model the rebrigthening at about 0.6 d with a simplestep function. We also tried more complex models by using apulse following the functional form discussed in Norris et al.(1996) and in Guidorzi et al. (2007a). The quality of data didnot allow us to study the transition in detail; we report the c (cid:13)000
BAT data were also analyzed. The 15–150 keV mask-weighted energy spectrum was extracted withthe ftool batbinevt using the BAT refined position (Krimmet al. 2007). All the required corrections were applied: weaccounted for the slewing through the ftool batupdatephakw and produced the detector-response matrix with batdrmgen .The spectrum was corrected for systematics depending onenergy with the ftool batphasyserr and was finally groupedby imposing a 3 σ threshold on each grouped energy channel. The optical/NIR light curve is complex, with an initialrising phase, a maximum, then a decay interrupted by arather sharp rebrightening, followed by a final steepening.We choose to fit the light curve with a Beuermann function(Eq. 1, Beuermann et al. 1999) for the first part down to therebrightening: F ( t ) = A/ [( t/t b ) κα r + ( t/t b ) κα d ] /κ , (1)where A is a normalisation constant, α r(d) is the slope ofthe rise (decay) phase, and κ is a smoothness parameter.The time at which the curve reaches its maximum is t max = t b ( − α r /α d ) / [ κ ( α d − α r )] .We model the rebrigthening at about 0.6 d with a simplestep function. We also tried more complex models by using apulse following the functional form discussed in Norris et al.(1996) and in Guidorzi et al. (2007a). The quality of data didnot allow us to study the transition in detail; we report the c (cid:13)000 , 1–11 S. Covino et al.
Table 1.
Optical/NIR observations of GRB 071010A. The reference time t GRB is Oct. 10, 2007 at 03:41:12 (Moretti et al. 2007). Dataare not corrected for dust absorption. For the TAROT data, the “clear” magnitudes are calibrated against the R filter. Data are sortedaccording to wavelength and according to time for each filter. Mean date t − t GRB Exp time Airmass Filter Instrument Magnitude(UT) (day) (s)2007 Oct 11.03601 0.88240 60 × U VLT+FORS2 20 . ± . × U VLT+FORS2 22 . ± . × B VLT+FORS2 21 . ± . × B VLT+FORS2 22 . ± . × V NOT+PolCor 20 . ± . × V VLT+FORS1 20 . ± . × V VLT+FORS1 20 . ± . × V VLT+FORS2 20 . ± . × V VLT+FORS1 20 . ± . × V VLT+FORS1 20 . ± . × V VLT+FORS1 20 . ± . × V VLT+FORS1 20 . ± . × V VLT+FORS1 20 . ± . × V VLT+FORS1 20 . ± . × V VLT+FORS1 20 . ± . × V Keck I+LRIS 20 . ± . × V Keck I+LRIS 20 . ± . × V VLT+FORS2 21 . ± . × V VLT+FORS2 22 . ± . × R Keck I+LRIS 18 . ± . × R Keck I+LRIS 18 . ± . × R Keck I+LRIS 18 . ± . × R Keck I+LRIS 18 . ± . × R Keck I+LRIS 18 . ± . × R Keck I+LRIS 18 . ± . × R Keck I+LRIS 19 . ± . × R Sampurnanand 20 . ± . × R VLT+FORS2 19 . ± . × R VLT+FORS1 19 . ± . × R Keck I+LRIS 20 . ± . × R Keck I+LRIS 20 . ± . × R Keck I+LRIS 20 . ± . × R Keck I+LRIS 20 . ± . × R VLT+FORS2 21 . ± . × R VLT+FORS2 22 . ± . × R VLT+FORS2 22 . ± . ×
15 1.2 R VLT+FORS2 23 . ± . ×
150 1.7 R Keck I+LRIS 24 . ± . × I NOT+PolCor 18 . ± . × I VLT+FORS2 19 . ± . × I VLT+FORS2 20 . ± . × I VLT+FORS2 22 . ± . ×
15 1.1 I VLT+FORS2 23 . ± . × I VLT+FORS2 24 . ± . × I VLT+FORS2 24 . ± . × J REM+REMIR 14 . ± . × J REM+REMIR 15 . ± . J GROND 15 . ± . × J REM+REMIR 15 . ± . × J GEMINI N+NIRI 16 . ± . J GROND 17 . ± . ×
15 1.1 J NTT+SofI 19 . ± . × H REM+REMIR 14 . ± . × H REM+REMIR 14 . ± . H GROND 14 . ± . × H REM+REMIR 14 . ± . ×
12 2.1 H GEMINI N+NIRI 16 . ± . ×
30 2.1 H TNG+NICS 16 . ± . H GROND 16 . ± . ×
20 1.2 H NTT+SofI 19 . ± . × K REM+REMIR 13 . ± . × K REM+REMIR 13 . ± . K GROND 13 . ± . × K REM+REMIR 14 . ± . × K GEMINI N+NIRI 15 . ± . K GROND 16 . ± . ×
25 1.1 K NTT+SofI 18 . ± . . ± . . ± . . ± . . ± . R TAROT 16 . ± . . ± . . ± . R TAROT 16 . ± . . ± . . ± . R TAROT 17 . ± . . ± . . ± . R TAROT 17 . ± . . ± . . ± . results requiring the minimum number of free parameters.A general discussion is given in Sec. 4.The later-time evolution of the afterglow shows a finalsteepening, and this was also modelled with a Beuermannfunction (Eq. 1). Both for the onset and the final steepen-ing the transition from the first to the second power-law be-haviour was remarkably sharp and the Beuermann’s smooth-ness parameter had a high value ( κ (cid:62)
10, frozen in the fits), making the model essentially consistent with what would beobtained by using simple power-law segments. The tempo-ral decay shown by the
Swift -XRT data was modelled withthe same functional form as the optical/NIR where quasi-simultaneous data were available.The data were fitted both in the spectral and tempo-ral domains. The optical/NIR afterglow spectrum was mod-elled with a simple power law. Galactic dust absorption was c (cid:13) , 1–11 he complex light-curve of the afterglow of GRB 071010A Figure 1.
The field of GRB 071010A at about 3.8 d after the GRBin the R band observed with the VLT equipped with FORS 2. Anearby source at R = 23 . ± .
06 at about one arcsec Eastwardof the afterglow is visible (see Sect. 3). The seeing of the image is ∼ . removed and rest-frame absorption was modelled with theabsorption curves of the Milky Way (MW), the Large andSmall Magellanic Clouds (LMC, SMC) (Pei 1992), and fora starburst galaxy (SB) (Calzetti et al. 2000). For the I band we also added a constant component, possibly the hostgalaxy or a supernova (SN), as late-time data (later thanabout 10 d) show a clear flattening. However, the presenceof a red point-like object ∼ (cid:48)(cid:48) east of the afterglow, with R = 23 . ± .
06 mag and I = 22 . ± .
07 mag, makeslate-time photometry less reliable (see Fig. 1). This sourcecould also contribute to the NIR flux at about 2.8 d, whenthe observing conditions did not allow a reliable separation.A simple extrapolation to the NIR bands for R − K ≈ χ = 114 . / ≈ . t max =420 +124 − s ( ∼ α r = − . +0 . − . , and α d = 0 . +0 . − . ,for the time of the maximum and the rising and decay-ing power-law indices, respectively, in substantial agreementwith the findings reported by Klotz et al. (2007b). The datado not suggest the presence of any spectral evolution.A rebrightening peaking at about 0 . ∼ .
9) is remarkably high and, within the uncer-
Table 2.
Best-fitting parameters.Parameter Interval α r − . +0 . − . t max +124 − s α d . +0 . − . t break . +0 . − . d α d . +0 . − . f inj /f ∼ . β opt / NIR . +0 . − . E B − V . +0 . − . β X . +0 . − . N H . +0 . − . ) × cm − tainties, the same in the optical/NIR and X-rays. The laterevolution of the afterglow requires a break, consistent withbeing fully achromatic in the optical/NIR and X-ray bands,at t break = 0 . +0 . − . d, with a post-break decay index of α d = 2 . +0 . − . , followed by a final flattening likely due tothe contribution from the GRB host galaxy. The best-fittingmodel is shown superposed on the data in Fig. 2.The entire afterglow spectral evolution is consistentwith being achromatic. In the optical/NIR the broad-bandspectral energy distribution index is β = 0 . +0 . − . , with asizeable rest-frame extinction E B − V = 0 . +0 . − . mag afterhaving removed the Milky Way absorption along the line ofsight. Among the various extinction curves, only that of theSMC gave satisfactory results. The redshift of GRB 071010Awould move the 2175 ˚A bump (see Fynbo et al. 2007, for adiscussion about observations of this feature in GRB after-glow spectra), typical of Milky Way absorption (Pei 1992),into the B band, therefore easily detectable, if present, inour data sets.The X-ray spectrum was modelled assuming a simplepower law with neutral absorption in the Milky Way andin the host galaxy. Analysis of the hardness ratio showedthat there are no significant spectral variations during theX-ray observations. The spectral analysis of the X-ray dataalone ( χ = 14 . / ≈ .
76) provided satisfactory results.Fixing the Galactic absorption at 6 . × cm − (Dickey &Lockman 1990) the rest-frame absorption and the spectralindex turned out to be 1 . +0 . − . ) × cm − and β =1 . +0 . − . , respectively. The optical/NIR and X-ray spectralindices are consistent within the uncertainties with β X ≈ β opt +0 .
5, i.e. with the cooling-break between the two bands.The broad-band SED is shown in Fig. 3.The prompt time-averaged spectrum of GRB 071010Ain the
Swift
BAT 15–150 keV energy band is consistent witha single power law with photon index significantly higherthan 2 (Krimm et al. 2007), suggesting that the peak en-ergy, E peak , of this event is close to, or below, 20–30 keV. Inorder to better investigate this issue, we reduced and ana-lyzed the BAT time-averaged spectrum. The fit with a sim-ple power law yelds a photon index of 3.1 ± α fixed at −
1, we derive an upper limit to the peak en-ergy E peak of ∼
35 keV, which corresponds to ∼
70 keV in theGRB cosmological rest frame. Then, by varying E peak fromits upper limit to 0 and adopting the method of Ghirlanda c (cid:13)000
70 keV in theGRB cosmological rest frame. Then, by varying E peak fromits upper limit to 0 and adopting the method of Ghirlanda c (cid:13)000 , 1–11 S. Covino et al. et al. (2004) and Amati (2006), we estimated an isotropic-equivalent radiated energy in the 1–10000 keV cosmologicalrest-frame energy band, E iso , of (3 . ± . × erg. The temporal resolution of the observations does not allowus to check in detail the early-time density profile of the cir-cumburst medium or possible chromaticity in the evolution.The modest duration of the prompt event and the late onsettime argues against a scenario with frequent multiple peaksas for GRB 070311 (Guidorzi et al. 2007a). A smooth risingphase can also be produced by a decreasing extinction in thecase of a radially decreasing circumburst density (Rykoff etal. 2004). However, the detection of the afterglow in the U band (Table 1) eliminates this last possibility due to thehigh absorption required. An alternative giving a chromaticmaximum could be the passage of the typical synchrotronfrequency in the optical/NIR bands. The time spread for themaximum would depend on the width of the observed fre-quency band, t /t ∝ ( ν /ν ) − / , with the higher frequen-cies peaking first. The passage would also produce a spectralchange from a positive (1 /
3) to negative [ − ( p − /
2] spectralindex in the slow-cooling case (Sari et al. 1998; Chevalier &Li 1999). Again, this is not supported by the data.The maximum during the early afterglow evolution canthen be interpreted as the afterglow onset as for the case ofGRB 060418 and GRB 060607A (Molinari et al. 2007; Jin &Fan 2007). Following the formalism reported in Eqs. 1 and2 of Molinari et al. (2007) in the thin-shell case since theprompt duration was much shorter than the peak of theafterglow emssison, we can derive, for GRB 071010A, Γ ≈
150 for a uniform interstellar medium (ISM) and Γ ≈
40 inthe case of a wind-shaped density profile. These are lowerthan the values inferred for GRB 060418 and GRB 060607Adue to the late occurrence of the maximum, but still withinthe theoretical expectations for the external-shock scenario(Sari & Piran 1999; Lithwick & Sari 2001; Zhang et al. 2006).The observed maximum cannot be related to reverseshock emission. The reverse shock should be in generala short-lasting phenomenon, and in the thin-shell case itshould peak slightly before the deceleration time, roughlyjust before the afterglow onset, and dominate it in case thetypical synchrotron emission is close enough to the opticalband (Kobayashi & Zhang 2007; Jin & Fan 2007). The rapiddecay ( f ∼ t − . , Sari & Piran 1999; Kobayashi & Zhang2007) predicted for the reverse shock emission is inconsistentwith the much milder decay observed after the optical fluxpeak (Table 2). The occurrence of rebrightenings or flares during the tem-poral evolution of GRB afterglows is not uncommon. Thelong time coverage usually provided by the
Swift
XRT af-fords the study of a large sample of flares from the spectraland energetic (Falcone et al. 2007), or temporal and mor-phological (Chincarini et al. 2007) points of view. In theoptical/NIR the occurrence of large flares seems to be less common, while in a remarkable fraction of cases a rebright-ening due to discrete energy injection was included in themodelling (J´ohannesson et al. 2006).Our dataset does not permit a detailed analysis of thetemporal and spectral evolution of the rebrightening around0.6 d. The simplest approach is to model the transition asdue to the discrete injection of a sizeable amount of en-ergy in the fireball interacting with the external medium bymeans of refreshed shocks. The flux density produced by anexternal shock scales with the energy content of the fireballapart from minor differences between the optical/NIR andX-ray bands (Panaitescu & Kumar 2000) which are not de-tectable in our data. Therefore, to account for the observed f inj /f ≈ .
9, an amount of energy comparable to the energycontent of the fireball needs to be supplied to the system.Remarkably, a discrete episode of energy injection is notsupposed to modify the spectrum of the synchrotron radia-tion emitted at the shock front if the cooling frequency doesnot enter or cross the observing bands (Panaitescu & Kumar2000), in agreement with the achromatic evolution of the af-terglow of GRB 071010A. A discrete energy injection, com-parable to the energy content of the fireball, was also sug-gested by de Ugarte Postigo et al. (2007) for GRB 050408.Also in the case of the XRF 050824 a rebrightening was ob-served at rather late time and a large delayed energy in-jection was invoked (Sollerman et al. 2007). In principle, arebrightening could be induced by a density variation in thecircumburst medium as proposed by Lazzati et al. (2002) forthe GRB 021004 (but see also Nakar & Granot 2007). How-ever, this explanation cannot be applied to our case, sincethe X-ray range is above the cooling frequency and there-fore insensitive to density variations (see, e.g., Sari et al.1998). A transition from wind to homogenous medium, dueto the crossing of the wind reverse shock could also generatea rebrightening, as discussed in Pe’er & Wijers (2006). Otherexplanations are also possible, such as a double jet (Berger etal. 2003) or a patchy shell (Rees & M´esz´aros 1998; M´esz´aroset al. 1998). A structured jet with a bright narrow core andfainter wings can also produce a rebrightening close to thejet-break time (Salmonson 2003), though in such a case itwould only be due to perspective rather than to real energyinjection, possibly driving to a sharp transition if seen froma suitable angle.Finally, although not required by the data, the temporalprofile of the rebrightening is also consistent with that of afast rise and exponential decay pulse simply superposed onthe underlying afterglow. The relatively long duration of thepulse, as compared to the time of occurrence, ∆ t/t ≈
1, iscompatible with an origin of the emission at the external-shock radius (see discussion in Guetta et al. 2007; Guidorziet al. 2007a).
Following the rebrightening at about 0.6 d, the afterglowevolution shows a sharp break at about 1 d consistentwith being achromatic in the X-ray and optical/NIR bands(Fig. 2). Following Liang et al. (2007) only seven GRB af-terglows have been found so far with such an achromatictransition. Moreover, the general temporal behaviour of theoptical/NIR and X-ray light curves seems to be the same, c (cid:13) , 1–11 he complex light-curve of the afterglow of GRB 071010A Figure 2.
The optical, NIR, and X-ray light curves of the afterglow of GRB 071010A. The dashed lines show the fit to the temporaldecay of the early and late-time afterglow. Fit parameters are reported in Sect. 3. On the left we show the whole light curve fitted withthe adopted model. On the right the epoch around 1 d is magnified. At late times, the long-dashed line shows the possible contributionto the I band by a supernova component using SN 1998bw (Galama et al. 1998) as a template, at the redshift of GRB 071010A. Theeffect of the absorption in the optical is included. supporting at least qualitatively the possibility that we areobserving emission from a single outflow with a transitiondue to a beamed geometry.The standard external-shock model predicts specific re-lations between temporal and spectral parameters. However,it is a well-established observational fact that GRB after-glows show more complex behaviours than expected beforethe launch of Swift . Many mechanisms have been proposedto interpret the new features discovered with
Swift : de-layed energy injection, late central engine activity, refreshedshocks, high-latitude emission, etc. (e.g., Zhang & M´esz´aros2004; Zhang et al. 2006). Most of these ingredients affect thelight curve shape in a number of ways. However, since thefundamental emission process in the soft X-rays and in theoptical/NIR is still synchrotron, the spectra should followthe general predictions of the external-shock scenario. Wetherefore begin the analysis starting from the broad-bandSED from the optical/NIR to the X-rays. The evolution ofthe afterglow, including the early phases, is consistent withbeing fully achromatic. We derive the SED at about 0.8 d inorder to minimise the need to extrapolate the observationsadopting a temporal behaviour (Fig. 3).The amount of rest-frame neutral absorption requiredby the X-ray data ( N H ≈ . × cm − ) is within therange observed for a set of well-studied Swift
GRB X-rayafterglows (Campana et al. 2006). In the optical/NIR theobserved SED curvature clearly indicates the need for a sub-stantial absorption, as derived in Sect. 3, with E B − V ≈ . Figure 3.
SED of the afterglow of GRB 071010A about 0.8 dafter the burst. The dashed line is a fit to the observed data inthe optical/NIR. The solid line refers to the intrinsic SED once theeffect of dust absorption in the Milky Way and in the GRB hostgalaxy is removed. The X-ray and optical/NIR broad-band SED isconsistent with the predictions of the synchrotron external-shockscenario with a cooling break between the optical/NIR and theX-ray and β X = β opt + 1 / (cid:13) , 1–11 S. Covino et al. cal afterglows, a SMC-like extinction curve seems to providea better fit to the data (Kann et al. 2006, 2007), althoughthe gas-to-dust ratio N H /E B − V is higher than in the SMC,again in agreement with the findings of Starling et al. (2007).The spectral indices in the optical/NIR and X-raybands allows us to constrain the electron energy distributionindex p ( N ∝ γ − p ). With the cooling frequency in betweenthe optical/NIR and the X-rays (probably just below the X-ray range, as shown in Fig. 3), we have β opt = ( p − / β X = p/ p opt = 2 . +0 . − . and p X = 2 . +0 . − . ,independent of the circumburst matter-density profile in theslow-cooling regime (Zhang & M´esz´aros 2004, and referencestherein). These values for p , in particular from the X-rays,lit in the high tail of the distribution derived by Tagliaferriet al. (2006) or Shen et al. (2006), although still consistentwith it. The value derived from the optical is also consistentwith the typical results ( p ≈ .
2) of numerical simulationsof relativistic shocks (Achterberg et al. 2001; Vietri 2003).Knowing the electron distribution index we can predict thedecay rate in the optical/NIR before the occurrence of ajet break, considering also the constraints from the X-rays: α pre , opt ≈ . − . α pre , opt ≈ . − . α pre , X ≈ . − .
75 (1.4-2.0from the X-ray only) independent of the density profile (andunconstrained by our data).It is clear from the fit results, reported in Sect. 3 andshown in Fig. 2, that interpreting the break at ∼ p value derived from the optical.Otherwise, a continuous additional injection of energy as re-quired by the modeling of numerous Swift
X-ray afterglows(Panaitescu et al. 2006) should be considered. After the jetbreak, the decay index in the slow-cooling regime must bethe same at all frequencies (higher than the typical syn-chrotron frequency, as expected ∼ p >
2, as in our case, it is simply α post = p . Againthe observed decay indices are consistent with the expecta-tions only with the value derived from the optical/NIR.In principle, the post-break decay indices would alsobe consistent with a wind model before the occurrence ofa jet break with p ≈ .
1. This p value would be bettersupported by the X-ray spectral data, although the achro-maticity of the transition at ∼ p values, although likely incon-sistent with a unique p ≈ . p < I band about 0.8 magfainter than our last measured point. It is well known that the position of a GRB in the radiatedenergy vs. spectral peak photon energy plane can provideuseful clues on the nature, physical origin, and geometryof its emission. Indeed, while all “normal” long GRBs andXRFs show a clear correlation between these two quanti-ties, short GRBs and peculiar sub-energetic GRBs (980425and, possibly, 031203) show a different behavior (e.g., Am-ati 2006). In addition, collimated GRBs seen off-axis are ex-pected to deviate from these spectrum-energy correlations.The E peak and E iso values derived from the analysisof the time-averaged spectrum (see Sect. 3) are fully con-sistent with the E p , i − E iso (Amati) relation (Amati et al.2002; Amati 2006), showing that the prompt emission ofGRB 071010A is not peculiar and indicating that, if theemission is collimated, the off-axis angle is small. In addi-tion, with a break interpreted as a jet break at about 1 d andby assuming, following Ghirlanda et al. (2004), a constantcircumburst matter-density profile ( n = 3 cm − ) and a fire-ball kinetic to radiated energy conversion efficiency of 0.2, wederive a jet opening angle of 8 . ± . ± × erg. Thus, GRB 071010A isalso fully consistent with the E p , i − E γ (Ghirlanda) relation(Ghirlanda et al. 2004; Nava et al. 2006). This evidence fur-ther supports that the possibly achromatic break observedin the afterglow light curve is due to collimated emission.Finally, if we fit the time-averaged spectrum ofGRB 071010A with a cut-off power law with index fixed at −
1, we can constrain the cosmological rest-frame peak en-ergy to 32 +46 − keV. This value, combined with the values of E iso and E γ derived above, make the representative pointfor GRB 071010A consistent with the best-fit power law ofboth the E p , i − E iso and E p , i − E γ relations, thus reinforcingthe above considerations. Almost coincident with the achromatic break, we obtainedpolarimetric observations, providing a rather robust upperlimit (0.9% at 95%, 1.3% at 3 σ ) suggesting null polarisa-tion. Polarimetric measurements of GRB optical afterglowshave been performed for several events (Covino et al. 2004).In general, the polarisation level was moderate (2–3%), al-though in a few cases evidence for variations related to theafterglow evolution was found (Lazzati et al. 2003; Bersieret al. 2003; Rol et al. 2003; Greiner et al. 2003; Gorosabelet al. 2004; Lazzati et al. 2004). The simple detection of c (cid:13) , 1–11 he complex light-curve of the afterglow of GRB 071010A variable polarization, intrinsically related to the afterglowemission, is still one of the most important observationaltests in favour of the external relativistic shock model withphysical beaming, where some degree of polarisation is nat-urally expected (Malesani et al. 2005; Covino et al. 2005;Lazzati 2006).Almost all of the observations have been performed withthe goal of disentangling the geometry of the outflow usingtime-resolved polarimetry. After an initial period of enthusi-asm, the required intense and delicate observational effortsand the high degree of complexity of the afterglow lightcurves provided by Swift cast some doubts on the actualpossibility of deriving meaningful constraints from simplemodels (Ghisellini & Lazzati 1999; Sari 1999). These provedto be applicable only to smooth, regular, afterglow lightcurves. Nevertheless, Rossi et al. (2004) showed that thepolarisation degree and position-angle evolution is stronglydependent on the assumed jet structure. At least in a cou-ple of cases a homogeneous jet geometry could be formallyruled out: i.e. for GRB 020813 (Lazzati et al. 2004) and forGRB 030328 (Maiorano et al. 2006).The diagnostic power of time-resolved polarimetric ob-servations relies on the strong coupling between the assumedjet geometry and polarization signature. For homogeneousjets there should be two polarization maxima, before andafter the jet-break time, with a rotation of the position an-gle by 90 ◦ . For structured jets, (i.e., jets with a more ener-getic core and fainter wings), the maximum of polarizationshould almost be coincident with the jet-break time, and thepolarization angle constant. Therefore, a polarimetric mea-surement almost coincident with a jet break, as in the casefor GRB 071010A, is of remarkable value. The very low up-per limit that we have obtained (Sect. 3) argues against astructured jet. However, the lack of time-resolved polarime-try through the evolution of the afterglow prevents us fromdrawing firm conclusions. The polarised fraction for the ra-diation emitted by an afterglow depends basically on thedegree of asymmetry of the system, which in turn dependson the line of sight of the observer with respect to the jetsymmetry axis. With a single observation the possibility of avery small offset angle cannot be ruled out. Setting aside thejet-break hypothesis, it is still true that, whichever physicalprocess is responsible for the (achromatic) steepening ob-served in the afterglow at about 1 d, it generated radiationwith only a very low level of polarization. The same is true ifwe model the rebrightening at about 0.6 d with a fast rise ex-ponential decline pulse dominating the afterglow luminosityfor the pulse duration. Our campaign on GRB 071010A is an example of extensivemonitoring of a GRB afterglow in the optical/NIR. Thisallowed us to detect an initial rising phase which can beinterpreted as the afterglow onset. The peak time is ratherlate, about 7 min after the GRB. This would directly implya relatively small initial Lorentz factor although still withinthe fireball model expectations. Interesting features are therebrightening at ∼ . ∼ Swift era.The spectral and temporal relations before and afterthe break are roughly consistent with the predictions of theafterglow synchrotron external shock model with a cooling-break in between the optical/NIR and X-ray bands and as-suming also continuous energy injection. The rebrighteningepisode can be interpreted as being due to the occurrence ofa refreshed shock injecting energy in the system comparablyto the original energy content of the fireball. Other expla-nations are still possible, relating the rebrightening to thedensity profile of the circumburst medium or to structuredjets. The optical/NIR spectral energy distribution requiresabsorption in the host-galaxy satisfactorily modeled withthe SMC extinction curve.GRB 071010A is consistent with the Amati relation and,if the achromatic steepening is interpreted as a jet break, itis consistent with the Ghirlanda relation too. Independentof the emission process, the entire afterglow evolution is con-sistent with being independent of color from about 2 min upto a few days. The SED is consistent with synchrotron emis-sion and requires rest-frame absorption both in the X-raysand in the optical/NIR.We also performed a polarimetric observation almost incoincidence with the occurrence of the achromatic break. Atight (0.9%) upper limit was derived. In the absence of time-resolved polarimetry, we cannot rule out that this afterglowwas characterized by a low level of polarization throughoutits evolution.Finally, the rich dataset of this burst and the simul-taneous availability of multi-wavelength data is a positiveexample of a new approach we are trying to apply to GRBslocalized by
Swift , favouring more complete coverage of a fewwell-studied GRBs. In addition, the observation of an initialrising phase, a rebrightening episode, and an achromaticbreak, as well as the availability of a well-sampled broad-band energy distribution, could offer a valuable workbenchfor alternative interpretative scenarios such as the “cannon-ball” (De R`ujula 2007; Dado et al. 2007), the “fireshell”(Bianco et al. 2007), and the “quark nova” (Staff et al. 2007)models in order to reach a better understanding of GRB af-terglow phenomenology.
ACKNOWLEDGMENTS
SC thanks G. Ghirlanda, G. Ghisellini, D. Malesani and F.Tavecchio for useful discussions. AVF’s group at UC Berke-ley is funded by NSF grant AST–0607485 and NASA/
Swift grant NNG06GI86G. BG acknowledges a post-doctoralgrant funded by the Centre National d’Etudes Spatiales. Wealso thank the anonymous referee for her/his useful com-ments and suggestions.
REFERENCES
Achterberg A., Gallant Y.A., Kirk J.G., Guthmann A.W.,2001, MNRAS, 328, 393Amati L., 2006, MNRAS, 372, 233Amati L., Frontera F., Tavani M. et al., 2002, A&A, 390,81 c (cid:13)000
Achterberg A., Gallant Y.A., Kirk J.G., Guthmann A.W.,2001, MNRAS, 328, 393Amati L., 2006, MNRAS, 372, 233Amati L., Frontera F., Tavani M. et al., 2002, A&A, 390,81 c (cid:13)000 , 1–11 S. Covino et al.
Band D., Matteson J., Ford L. et al., 1993, ApJ, 413, 281Berger E., Kulkarni S.R., Fox D.B. et al., 2006, ApJ, 634,501Berger E., Kulkarni S.R., Pooley G. et al., 2003, Nature,426, 154Bersier D., McLeod B., Garnavich P. et al., 2003, ApJ, 583,63Bertin E., Arnouts S., 1995, A&AS, 177, 393Beuermann K., Hessman F.V., Reinsch K. et al., 1999,A&A, 352, 26Bianco C.L., Bernardini M.G., Caito L. et al., 2007, AIPConf. Procf., 966, 12 (astro-ph/0712.0495)Bloom J.S., Perley D.A., Modjaz M. et al., 2007, GCN 6861Bringer M., Bo¨er M., Peignot C., Fontan G., Merce C. 1999,A&AS, 138, 581Burrows D.N., Racusin J., 2006, NCim B, 121, 1273Calzetti D., Armus L., Bohlin R.C. et al., 2000, ApJ, 533,682Campana S., Romano P., Covino S. et al., 2006, A&A, 449,61Chandra P., Frail D.A., 2007, GCN 6901Chevalier R.A., Li Z., 1999, ApJ, 520, 29Chincarini G., Moretti A., Romano P. et al., 2007, ApJ,671, 1921Conconi P., Cunniffe R., D’Alessio F., et al., 2004, SPIE,5492, 1602Covino S., Ghisellini G., Lazzati D., Malesani D., 2004,ASP Conf. Ser., 312, 169Covino S., Lazzati D., Ghisellini G. et al.,1999, A&A, 348,1Covino S., Lazzati D., Malesani D. et al., 2002, A&A, 392,865Covino S., Malesani D., Ghisellini G. et al., 2003, A&A,400, 9Covino S., Malesani D., Tagliaferri G. et al., 2006, NCimB, 121, 1171Covino S., Rossi E., Lazzati D., Malesani D., Ghisellini G.,2005, AIP Conf. Proc., 797, 144Curran P.A., van der Horst A.J., Wijers R.A.M.J., 2007,astro-ph/0710.5285Dado S., Dar A., De Rujula A. 2007, astro-ph/0712.1527Dai X., Garnavich P.M., Prieto J.L. et al., 2007, astro-ph/0712.2239De R`ujula A., 2007, astro-ph/0711.0970D’Elia V., Fiore F., Piranomonte S. et al., 2008, in prepa-rationde Ugarte Postigo A., Fatkhullin T.A., J´ohannesson et al.,2007, A&A, 462, 57Devillard N., 1997, The Messenger, 87Dickey J.M., Lockman F.J., 1990, ARAA, 28, 15Falcone A.D., Morris D., Racusin J. et al., 2007, ApJ, 671,1921Fynbo J., Vreeswijk P., Jakobsson P. et al., 2007, The Mes-senger, 130, 43Galama T.J.., Vreeswijk P.M., van Paradijs J. et al., 1998,Nature, 395, 670Gehrels N., Chincarini G., Giommi P. et al., 2004, ApJ,611, 1005Ghirlanda G., Ghisellini G., Lazzati D., 2004, ApJ, 616,331Ghisellini G., Lazzati D., 1999, MNRAS, 309, 7Gorosabel J., Rol E., Covino S. et al., 2004, A&A, 422, 113 Greiner J., Bornemann W., Clemens C. et al., 2008, PASP,120 (in press) astro-ph/0801.4801Greiner J., Klose S., Reinsch K. et al., 2003, Nature, 426,157Guetta D., Fiore F., D’Elia V. et al., 2007, A&A, 461, 95Guidorzi C., Mangano V., Moretti A. et al., 2007b,GCN 6870Guidorzi C., Vergani S.D., Sazonov S. et al., 2007a, A&A,474, 793Jakobsson P., Levan A., Fynbo J.P.U. et al., 2006, A&A,447, 897Jin Z.P., Fan Y.Z., 2007, MNRAS, 378, 1043J´ohannesson G., Bj¨ornsson G., Gudmundsson E.H., 2006,ApJ, 647, 1238Kann D.A., Klose S., Zhang B. et al., 2007, as-tro.ph/0712.2186Kann D.A., Klose S., Zeh A., 2006, ApJ, 641, 993Klotz A., B¨oer M., Atteia J.L. 2007a, GCN 6860Klotz A., B¨oer M., Atteia J.L. 2007b, GCN 6863Kobayashi S., Zhang B. 2007, ApJ, 655, 973Krimm H., Barthelmy S.D., Cummings J. et al., 2007,GCN 6868Lazzati D. 2006, NJPh, 8, 131Lazzati D., Covino S., di Serego Alighieri S. et al., 2003,A&A, 410, 823Lazzati D., Covino S., Gorosabel J. et al., 2004, A&A, 422,121Lazzati D., Rossi E., Covino S., Ghisellini G., Malesani D.,2002, A&A, 396, 5Liang E-W., Racusin J.L., Zhang B., Zhang B-B. 2007,astro-ph/0708.2942Lithwick Y., Sari R., 2001, ApJ, 555, 540Maiorano E., Masetti N., Palazzi E. et al., 2006, A&A, 455,423Malesani D., Covino S., Rossi E., 2005, NCim C, 28, 515M´esz´aros P., Rees M.J., Wijers R.A.M.J., 1998, ApJ, 499,301Molinari E., Vergani S.D., Malesani D. et al., 2007, A&A,469, 13Monet D.G., Levine S.E., Canzian B. et al., 2003, AJ, 125,984Moretti A., Baumgartner W.H., Gehrels N. et al., 2007,GCN 6859Nakar E., Granot J., 2007, MNRAS, 380, 1744Nava L., Ghisellini G., Ghirlanda G., Tavecchio F., FirmaniC., 2006, A&A, 450, 471Norris J.P., Nemiroff R.J., Bonnell J.T. 1996, ApJ, 459,393Nousek J.A., Kouveliotou C., Grupe D. et al., 2006, ApJ,642, 389Panaitescu A., 2008, MNRAS, 383, 1143Panaitescu A., Kumar P., 2000, ApJ, 543, 66Panaitescu A., M´esz´aros P., Gehrels N., Burrows D.,Nousek J., 2006, MNRAS, 366, 1357Pe’er A., Wijers R.A.M.J. 2006, ApJ, 643, 1036Pei Y.C., 1992, ApJ, 395, 130Perley D.A., Filippenko A.V., Silverman J.M. et al., 2007,GCN 6934Prochaska J.X., Perley D.A., Modjaz M. et al., 2007,GCN 6864Rees M.J., M´esz´aros P., 1998, ApJ, 496, 1Rol E., Wijers R.A.M.J., Fynbo J.P.U. et al., 2003, A&A, c (cid:13) , 1–11 he complex light-curve of the afterglow of GRB 071010A c (cid:13)000