arXiv:astro-ph/0306137v2 1 Aug 2003 te yeo h fego aiblt sas observed also is variability afterglow the Frail in of references type e.g. Other (see, sources GRB ultrarelativis break in collimated These of wavelengths bursts. signatures combinations all to the attributed after their are at times or different curves laws at breaks power with as the approximated Usually be can date. to tail nywe ai n cuaeGBlclztoswr ob- with were (first localizations X-rays GRB in accurate tained and rapid possible obser- when become have The only afterglows optical bursts. dis- GRB gamma-ray the cosmological of of vations sources the the establishing to GRBs scale their tance of on number redshifts the information a measure to for important allowed they very example, For a observations sources. optical obtain gamma, to and X-rays allowed hard in emitted is aeegh rmrdot am swl eg Taylor (e.g. other Costa well in 1998, as found gamma were to afterglows radio from GRB wavelengths The (1997). Bond by al. et AHSLCS15.05.03 PAZHASTL.CLS Transla 573–578. p. 9, No. 29, Vol. 2003, Letters, Astronomy * ubro h pia fegoswr bevdi de- in observed were afterglows optical the of number A lhuhtemi ato h nryo am-a bursts gamma-ray of energy the of part main the Although -al [email protected] e-mail: 97) h rtotclatrlwwsdiscovered was afterglow optical first The 1997a).

c otepwrlwsetu ihseta index spectral with spectrum law power the to aigwt h pe iis1–%o iesae 0 scales time on 10–1% limits upper the with fading hspwrlwbekocr at occurs break law power This ih uv.Tepwrlwdcyidxsotl prahst approaches smoothly index filters decay BVRI law all in power that, The After curve. filters. light Bessel BVRI the of each us.Drn h rt5huso u bevtosteafterg the observations our of hours aftergl 5 the first to the pointed During telescopes burst. 1.5m-class first the of one ihte15 usa-uks eecp T10(T RTT150 telescope Russian-Turkish extrem the 1.5m of the observations with the of results first the present We Keywords: sp coll synchrotron with of RTT150. consistent signature is a x-rays as and optical interpreted radio, be can observations our of 08R .Burenin A. R. 2018 .Y.Tkachenko Yu. A. tal. et am-a uss—atrlw pia observations optical — afterglows — bursts gamma-ray 97,Burenin 1997b, IS OR FTEGB002 PIA AFTERGLOW OPTICAL 030329 GRB THE OF HOURS FIRST INTRODUCTION Beppo 1 *1 .Aslan Z. , .A Sunyaev A. R. , A aelt,eg Costa e.g. satellite, SAX tal. et t .Bikmaev I. − t 0 1999). 2 ≈ 1 T 2 5 pc eerhIsiue(K) ocw Russia Moscow, (IKI), Institute Research Space BTKNtoa bevtr,Atla Turkey Antalya, Observatory, National UBITAK .Khamitov I. , 0 aa tt nvriy(G) aa,Russia Kazan, (KGU), University State Kazan ¨ . 7dy n at for lasts and days 57 3 tal. et 5 aac nvriy sabl Turkey Istanbul, University, Sabanci .Sakhibullin N. , 1 α .N Pavlinsky N. M. , eevd2 pi 2003 April 28 Received 4 = i jets tic 2001). EU naa Turkey Ankara, METU, . –00s pcrlflxdsrbto nfu VIfitr corr filters BVRI four in distribution flux Spectral s. 1–1000 BTKNtoa bevtr,Bkryee uky.RTT15 Turkey). Bakyrlytepe, Observatory, National UBITAK tal. et e rmPsm srnmcekjZhurnal. Astronomicheskij v Pis’ma from ted 0 ¨ . 66 2 K.Uluc , crm ecniu u bevtoso hsuiu betw object unique this of observations our continue We ectrum. s ± 573 w bevtoswr tre prxmtl or fe t after hours 6 approximately started were Observations ow. l rgtotclatrlwo am-a us GB 03032 (GRB) burst gamma-ray of afterglow optical bright ely o ae xcl sapwrlwwt index with law power a as exactly faded low mtdutaeaiitcjt h fego u distribut flux afterglow The jet. ultrarelativistic imated 0 . 1 h hneo h oe a ea ne nteend the in index decay law power the of change The 01. nO pcrmltr(Stanek later spectrum OT in hf ecp o R 845 fi sascae ihsuper- with associated is at it 1998bw if nova 980425, GRB for (except shift al. et evtoso hseteeybih pia fego with afterglow optical bright extremely this of servations epnst iorpc nryrlaei am-aso or- of gamma-rays 10 in der release energy “isotropic” to responds hf fteO a eemndt be to determined was OT the of shift ain fti betaewl ouetdi C circulars GCN Burenin in documented (e.g. well are object obser- this spectroscopic the of and vations has throughout photometric object telescopes extensive this many The moment for world. target the prime At a 2003). become Torii (Peterson HETE-II 2003, by Price measured & that po with the consistent in discovered sition was (OT) transient optical bright very a 040kVeeg ag,addrto bu 0s(Van- s 30 about — derspek duration and range, energy keV 30–400 us ekflxwsmaue ob 7 be to measured was The flux UT. peak 11:37:14.7 burst at satellite HETE-II aboard instruments (e. supernovae 2001). underlying Sokolov of signatures as interpreted and ± ≈ iutnosyw bev teeigo h oe law power the of steepening a observe we simultaneously 5 evalue he nti ae epeettefis eut fteob- the of results first the present we paper this In eybih am-a us 339wsdtce by detected was 030329 burst gamma-ray bright very A .Suleymanov V. , 2 0 .A Alpar A. M. , 1 . 03,rnigi sacoetGBwt esrdred- measured with GRB closest a as it ranging 2003), as eosren aiblt bv h gradual the above variability no observe We days. 1 52 .V Denissenko V. D. , tal. et r.Teudryn uenv opnn emerged component underlying The erg. − ≈ 03 Golenetskii 2003, 1 tal. et . ,osre yohrosraoislater. observatories other by observed 9, z = 3 5 .Kiziloglu U. , 0 03 Khamitov 2003, . 05.Ti itnet h orecor- source the to distance This 0085). 1 .V Terekhov V. O. , tal. et tal. et 4 .Baykal A. , 2003). · tal. et 03.Wti hours 2 Within 2003). − 10 z 1 = − . 19 6 03.Tered- The 2003). 0 1 r s erg ± . , esponds 65(Greiner 1685 0 4 o in ion . was 0 , 1in 01 − 1 ith he 9 cm − 2 in g. - 574 BURENIN et al.

Fig. 1. The image of the GRB 030329 afterglow, obtained with RTT150 telescope on March 29 (left) and on May 28 (right). Arrows show the optical transient (OT) and the reference star (A) used to calibrate the flux of the OT (see text). the 1.5m Russian-Turkish telescope (TUBITAK¨ National shows the image of this field obtained on May 28. Observatory, Bakyrlytepe, Turkey, 2547m, 2h01m20s E, ◦ ′ ′′ 36 49 30 N). RTT150 turns out to be one of the first 1.5m- PHOTOMETRY class telescopes pointed to the afterglow. Observations were started already in approximately 6 hours after the burst. At All the measurements of the OT flux during the first night the beginning of our observations the magnitude of the op- of our observations were done relative to the nearby bright tical afterglow in R was estimated to be 14m. This is 3– star, designated as “A” in Fig. 1. This is the brightest star in 4m brighter than any GRB afterglow previously observed on our field of view. Nevertheless, during the first night even this time scale. the star “A” is much fainter than the OT. Thus the errors of the afterglow flux measurements during the first night are dominatedby the flux errorsfor this star. All the results were OBSERVATIONS checked using the other stars in the field, but all of them are We used an AP47p Apogee CCD mounted in the much fainter than star “A” and can not be used to improve Cassegrain focus of the telescope (1 : 7.7). This is a back- the OT photometry. illuminated 1056 × 1024 CCD chip. We used 2 × 2 binning To calibrate the field we observed the Landolt (1992) stars mode with 0.46′′ pixels. As always, bias and dark counts before and after the afterglow observations. These obser- were subtracted and flat field correction applied for all the vations show that night was almost perfectly photometric. images. The data reduction was done with IRAF (Image Small decrease of atmosphere transparency was observed Reduction and Analysis Facility)1 and using our own soft- during the very beginning of the night. The transparency ware. was lower by approximately 10% and returned to its usual At the very beginning of the night on 29 March 2003, value during the first few hours of the observations. at 17:50 UT, the telescope was pointed to the afterglow Our photometric calibration is in good agreement with the of gamma-ray burst 030329. We observed the object till photometry of this field provided by Henden (2003). In this 01:30 UT when its zenith distance had become higher than circular it was noted that the OT and star “A” have very dif- 68◦. All this time the OT field was imaged in BVRI Bessel ferent colors. Therefore using star “A” as a reference may filters in series of exposures 30 – 10 s (the readout time was cause an additional systematic error correlated with airmass about 10 s). In total about 700 images were taken, 175 in due to possible differential airmass corrections. However, each BVRI filter. The combined R image of the field is in our photometric solutions we found no variations of the shown in Fig. 1 in the left. The right panel of this figure color coefficients with airmass or with the decrease of the transparency during the first hours of our observations. We 1http://tucana.tuc.noao.edu/ estimate the systematic errors of the OT flux measurements

ASTRONOMY LETTERS Vol. 29 No. 9 2003 GRB030329OPTICALAFTERGLOW 575 t R 934 lg . m 2 − m

(GCN 2018)

t −t0, days t −t0, days

Fig. 2. The light curve of the GRB 030329 optical afterglow during the first night of the observations. The left panel shows the light curve in R band. The right panel gives the light curve in the BVRI bands after the subtraction of the power law decay with index −1.19 ± 0.01, obtained by fitting the first 5 hours of the R band light curve. The left panel also shows the power law with index −1.9, obtained at t −t0 > 0.6 days (Garnavich et al. 2003). In the right panel the triangles at t −t0 > 0.7 days show the data from Fitzgerald & Orosz (2003).

to be not more than 1%. t − t0 ≈ 0.57 days and lasted for approximately ±0.1 days. The light curve of the GRB 030329 afterglow in R band, Note that just before the break we observe some marginally obtained in the first night of our observations, is shown in significant flattening of the afterglow light curve in R and I. the left panel of Fig. 2. During the first 5 hours the afterglow Before our observations, at t −t0 < 0.25, there were unfil- flux decayed with good accuracy as a power law with index tered observations with small telescopes (e.g. Rykoff et al. −1.19 ± 0.01. After that for approximately 3 hours we ob- 2003, Sato et al. 2003). These data should be calibrated serve a steepening of the power law light curve. The power properly to be compared with our observations. We note law decay index smoothly approaches the value ≈−1.9, ob- however that the power law slope of the afterglow light served later by the observatories in the western hemisphere. curve at t −t0 from 0.06 till 0.17 days (from 1.5 till 4 hours) The left panel also shows the power law with index −1.9, equals to 1.07 ± 0.08 as can be inferred from the ROTSE obtained at t − t0 > 0.6 days with FLWO 1.2m telescope data (Rykoff et al. 2003), which is approximately consis- (Garnavich et al. 2003). tent with our measurements at later t −t0. An approximately The right panel of Fig. 2 shows the light curve in BVRI similar slope was observed also in (Nishihara et al. filters after the subtraction of power law decay with in- 2003). Thus, it is likely that the the power law decay index dex −1.19 ± 0.01, obtained by fitting the first 5 hours of of the afterglow does not change significantly from as early the R band light curve. We see that the power law decay as 1–2 hours after the burst. is the same in each of the BVRI filters. The power law slopes in BVI are equal to −1.22 ± 0.05, −1.19 ± 0.02 and −1.19 ± 0.01 respectively, which is within the errors equal SHORT TIME SCALE VARIABILITY to the power law slope in R. At t − t0 ≈ 0.5 days the flux In Fig. 2 one can see that we observe no variability of decay in all filters simultaneously starts to deviate from this GRB 030329 afterglowflux abovethe gradualdecay on time power law. Closer to the end of our observations we ob- scales 100–1000 s. Detailed analysis of the aperiodic vari- serve somewhat higher deviation in B. In the right panel of ablity of the afterglowlight curve is beyond the scope of this Fig. 2 the triangles at t − t0 > 0.7 days show the data from paper and will be a subject of the forthcoming publication. Fitzgerald & Orosz (2003). These data are approximately We note however, that RMS deviation of the observed light consistent with our measurements at t −t0 < 0.6 days. curve in R band from the best fit power law equals to ≈ 1% Comparing our data at t −t0 < 0.6 with the dataof western on these time scales. This number obviously includes con- observatories at t −t0 > 0.6 (see Fig. 2) one can see, that we tribution of the statistical and systematical errors and there- were lucky to observe in detail the beginning of the power fore gives a conservative estimate of the upper limit on the law break of the afterglow light curve. The break occurred at intrinsic variablity of the optical flux.

ASTRONOMY LETTERS Vol. 29 No. 9 2003 576 BURENIN et al. ADU , mJy

ν F

t, s

Fig. 3. A part of the afterglow light curve, obtained with high temporal resolution.

To examine the variability of the afterglow flux on even shorter time scales we observed the OT field, turning off the frequency, 1014 Hz hour tracking of the telescope. The OT was moving through the field of view during these exposures. We then recon- Fig. 4. Spectrum of the afterglow at t − t0 = 0.25 days (6 hours), in- structed the OT light curve with temporal resolution up to ferred from the fluxes in four BVRI bands, corrected for galactic extinc- 0.1s usingthe trackof the OTin CCD frame. In this method tion E(B −V)= 0.025 (Schlegel et al. 1998). This form of the spectrum remains exactly constant during the first 5 hours of our observations. one image contains a segment of the OT light curve of ap- proximately 15 s duration. Twenty images were taken in this way between 20:00 and 20:30 UT. One of those OT light of the afterglow power is approximately in X-rays. The curve segments is shown in Fig. 3. From these data we esti- maximum is very wide covering far and probably mate the RMS of the afterglow optical flux on 0.1–10 s time soft gamma-rays. Even in the optical the afterglow power is scales to be equal to 10–3% respectively, which again gives only a factor of two lower than the power in X-rays. conservative upper limits on the intrinsic variablity of the The first observation of this afterglow in the radio was afterglow. made approximately 0.6 days (14 hours) after the burst. The afterglow flux at 8.46GHz was measured to be 3.5mJy SPECTRUM ACROSS THE BVRI BANDS (Berger et al. 2003). Between t − t0 = 0.25 and 0.6 days In Fig. 2 one can see that the colors of the optical after- the afterglow radio flux probably had not changed by more glow of GRB 030329 are constant within the errors dur- than an order of magnitude. The afterglow radio flux was ing the first 5 hours of our observations. The low resolu- observed to be raising after the first observation (e.g. Poo- tion spectrum obtained from the fluxes in four BVRI filters ley 2003). Therefore, at t −t0 = 0.25 the afterglow power in radio should be approximately by 5–6 orders of magnitude at t − t = 0.25 days, and corrected for galactic extinction 0 lower than in optical or X-rays. E(B −V)= 0.025 (Schlegel et al. 1998) is shown in Fig. 4. With good accuracy it can be described as a power law with spectral index α = 0.66±0.01. We observea somewhat flat- DISCUSSION ter spectrum than that obtained by Stanek et al. (2003), who We present a high sensitivity observation of the light measured a power law spectral index α = 0.85, 2.6 days af- curve of the GRB 030329 optical afterglow, starting as early ter the burst. Note that both indices are not corrected for the as 6 hours after the burst. In each of the BVRI filters we extinction in GRB host galaxy. observe a gradual flux decay, which can be accurately de- In Fig. 5 we compare the afterglow fluxes in the optical, scribed as a power law Fν ∝ t−1.19 during the first 5 hours in X-rays and in radio. The RTT150 data are shown for of our observations. After that the afterglow flux started to t −t0 = 0.25 days (6 hours), when the second RXTE obser- decline faster. vation was made (Marshall & Swank 2003). We take the High signal-to-noise ratio during our observations al- X-ray flux and the power law spectral slope from this circu- lowed us to investigate the variability of the afterglow flux lar. According to this circular between t −t0 = 5 and6 hours with much higher temporal resolution than it could be done the X-ray flux decayed by 20%. This correspondsto a power previously. Within the errors of flux measurements we ob- law with slope approximately −1, which is close to what we serve no variability above the gradual fading down to the observe in the optical at about the same time. 0.1 s time scale. The upper limits are 10, 3 and 1% on 0.1, FromFig.5wesee thatat t −t0 = 0.25 days the maximum 10 and 100–1000 s time scales respectively.

ASTRONOMY LETTERS Vol. 29 No. 9 2003 GRB030329OPTICALAFTERGLOW 577

The observations of this unique object are continued now with RTT150. The results of these observations will be pre- sented in the subsequent papers.

We are grateful to Mikhail Revnivtsev for his assistance RXTE RTT150 (GCN 1996) in the analysis of RXTE data and to Sergey Sazonov for useful discussion of the results of our observations. This work was supported by Russian Fund for Basic Researches (grants 02-02-16619, 03-02-06768, 02-02-17342), by Rus-

, JyHz sian Government Program of Leading Science Schools Sup- ν F port (grant 2083.2003.2), by the program of Presidium of ν Russian Academy of Sciences “Variable phenomena in as- tronomy”. This work was also supported by the High Energy Astrophysics Working Group of the Scientific and VLA Technical Research Council of Turkey (TUBITAK)¨ through (GCN 2014) its support of basic research in Turkish universities and by the Turkish Academy of Sciences (for MAA).

frequency, Hz REFERENCES

Fig. 5. Afterglow νFν spectrum in radio, optical and X-ray bands. The RTT150 and RXTE data (Marshall & Swank 2003) are shown as was Berger E., Soderberg A. M., Frail D. A., GCN Circ. 2014 (2003) observed at t − t0 = 0.25 days (6 hours). The radio data correspond to Bond H. E. IAU Circ. 6654 (1997) t −t0 = 0.6 days (14 hours) were taken from Berger et al. (2003). Burenin R. A., Vikhlinin A. A., Gilfanov M. R., Terekhov O. V., Tkachenko A. Yu., Sazonov S. Yu., Churazov E. M., Sunyaev R. A., Goldoni P., Claret A., Goldwurm A., Paul J., The results of our observations are consistent with the Roques J. P., Jourdain E., Pelaez F., Vedrenne G., Astron. model where the afterglow emission is generated during the Astrophys., 344, L53 (1999); (astro-ph/9902006) deceleration of the ultrarelativistic collimated jet (see, e.g. Burenin R., Sunyaev R., Pavlinsky M., Denissenko D., Hurley et al. 2003). The break in power law light curve, Terekhov O., Tkachenko A., Aslan Z., Uluc K., Khamitov I., which we observe at t −t0 = 0.57, can be interpreted as the Kiziloglu U., Alpar A., Baykal A., Bikmaev I., Sakhibullin N., “jet break”, i.e. the break which occurs when the angular Suleymanov V., GCN Circ. 2001, 2024, 2046, 2051, 2054, 2079 structure of the ultrarelativistic collimated jet becomes ob- (2003) servable. This interpretation is supported by the fact that to Costa E., Feroci M., Piro L., et al., IAU Circ. 6649 (1997a) good accuracy this break occurred simultaneously in differ- Costa E., Frontera F., Heise J., et al., Nature, 387, 783 (1997b) ent colors. The power law slope of the light curve changes Fitzgerald J. B., Orosz J. A., GCN Circ. 2056 (2003) from −1.19 to −1.9, approximately as it is usually observed Frail D. A., Kulkarni S. R., Sari R., et al., Astrophys. J., 562, L55 in jet breaks (e.g. Harrison et al. 1999). Furthermore, before (2001) jet break there can not be any short time scale variability — Golenetskii S., Mazets E., Pal’shin V., Frederiks D., Cline T., GCN exactly what we observe with good accuracy. The variability Circ. 2026 (2003) indeed occurs after this break (see GCN circulars). Garnavich P., Stanek K. Z., Berlind P., GCN Circ. 2018 (2003) Greiner J., Peimbert M., Estaban C., et al., GCN Circ. 2020 (2003) For a uniform jet moving toward the observer, the time Harrison F. A., Bloom J. S., Frail D. A, et al., Astrophys. J., 523, of the jet break corresponds to the time when the gamma- L121 (1999) factor of the jet falls below θ −1, the inverse opening angle Henden A., GCN Circ. 2023, 2082 (2003) of the jet. We can determine the opening angle θ = 0.08 Hurley K., Sari R., Djorgovski S. G., astro-ph/0211620 (2003) using the formula and typical parameters from Frail et al. Khamitov I., Aslan Z., Golbasi O., Parmaksizoglu M., Bikmaev I., (2001). The actual energy release in gamma-rays appears to Sakhibullin N., Suleymanov V., Burenin R., Sunyaev R., 52θ 2 49 be 10 /2 = 3·10 erg, approximately an order of mag- Denissenko D., Pavlinsky M., Terekhov O., Tkachenko A., nitude lower than the typical value obtained by Frail et al. Kiziloglu U., Alpar A., Baykal A., GCN Circ. 2094, 2105, 2108, (2001) and comparable to the energy emitted by a typical 2119 (2003) supernova. Landolt A., Astron. J., 104, 340 (1992) The form of the afterglow flux distribution in the radio, Marshall F. E., Swank J. H., GCN Circ. 1996 (2003) optical and X-ray bands (Fig. 5) is approximately consis- Peterson B. A., Price P. A., GCN Circ. 1985 (2003) tent with synchrotron emission. Here the synchrotron cool- Pooley G., GCN Circ. 2043 (2003) ing frequency could be between optical and X-rays. Syn- Rykoff E. S., Smith D. A., GCN Circ. 1995 (2003) chrotron self-absorption is probably effective in radio band. Sato R., Yatsu Y., Suzuki M., Kawai N., GCN Circ. 2080 (2003)

ASTRONOMY LETTERS Vol. 29 No. 9 2003 578 BURENIN et al.

Sokolov V. V., Bulletin of the Special Astrophysical Observatory, (Izvestiya Spetsial’noi Astrofizicheskoi Observatorii), 51, 38 (2001) Stanek K. Z., Matheson T., Garnavich P. M., et al., Astrophys. J. ,in press, (2003); (astro-ph/0304173) Schlegel D. J., Finkbeiner D. P., Davis M. Astrophys. J., 500, 525 (1998) Taylor G. B., Frail D. A., Kulkarni S. R., Shepherd D. S., Feroci M., Frontera F., Astrophys. J., 502, L115 (1998) Torii K., GCN Circ. 1986 (2003) Vanderspek R., Crew G., Doty J., et al., GCN Circ. 1997 (2003)

ASTRONOMY LETTERS Vol. 29 No. 9 2003