Astron. Astrophys. 358, L41-L44 (2000)

2. Observation of the early afterglow emissions in GRB 910402 and GRB 920723

During detection of these two gamma-ray bursts the GRANAT satellite was outside of the Earth radiation belts and the Earth magnetosphere. There were no solar flares during these gamma-ray bursts (Solar Geophys. Data). Detailed analysis of the gamma-ray light curves shows that the fluxes after the end of the main events are not decreased to the background level during several hundreds of seconds. The background subtracted time histories of these two bursts in energy range in logarithmic scale are presented in the Fig. 3 and 5.

Investigating behaviour of the time history after the end of the the main gamma-ray burst we found that in all 6 detectors the count rate is varying around decreasing as the power law of time average values with the standard deviation correspondent to the Poissonian distribution. The trend of the count rate after these two bursts can be described by the same law (1) for .

Using this simple model it is possible to define intensity of the afterglow emission I and power law index of time for each of six PHEBUS detectors. It was also possible to estimate - time when the afterglow emission fading as the power-law with time begins to dominate over the gamma-ray bursts emission.

We have found that the excessive (over background level in each of the six PHEBUS BGO detectors) count rates I were proportional to the intensities of the main GRB events in each detector as it had been expected in the case if the afterglow emission is really connected with gamma-ray burst source. For each of the bursts the parameter was equal within statistical errors for all detectors of the instrument.

We have found that for GRB 910402 the best fit corresponds to (at confidence level) and (). The confidence interval for is from 62.2 to 77.9 . To obtain these parameters we have used the afterglow count rate data from 85 to 700  after .

For GRB 920723 the best fit corresponds to (at confidence level) and (). The confidence interval for is from 3.5 to 7.3  To obtain these parameters we have used the afterglow count rate data from 8 to 700  after .

The energy spectra of the main bursts and afterglow emission are different. In both cases (GRB 910402 and GRB 920723) just after GRB event the energy spectrum of the afterglow emissions is much softer than the energy spectrum of the main GRB events. Figs. 4 and 6 show the evolution of the spectral hardness of the bursts. We define spectral hardness as the ratio of count rate in to count rate) We define the effective photon index as the index of a power-law spectrum (in the energy range) which, would produce the observed hardness at an incidence angle of 90^o to the detector axis.

At the end of the both bursts the spectral hardness ratio is equal to (Fig. 4 and 6). This corresponds to the photon index of at the same energy range. For the GRB 910402 during the time interval less than we observe drop of the hardness down to the value of (Fig. 4). This value corresponds to the photon index of . The moment of abrupt drop of the hardness corresponds to for GRB 910402 (Fig. 4). For the burst GRB 920723 the time of the abrupt drop of the spectral hardness down to corresponds to the moment (Fig. 6). Thus the times of abrupt softening of the burst spectra coincide within statistical errors with the () - beginning of the power law emission fading during afterglow for both GRB 910402 and GRB 920723.

Note that as one can see from Fig. 4 during the afterglow emission of the GRB 910402 a statistically significant hardening of the spectra is observed. In the case of GRB 920723 the error bars are too large (Fig. 6) to make any conclusions about behaviour of the afterglow spectra.

The afterglow emission intensities in energy range are rather faint in comparison with gamma-ray bursts. Just a few tens of seconds after gamma-ray bursts discussed in this paper the afterglow intensities are of the GRB maximum intensities. In the case of GRB 920723 the afterglow emission contains of the total burst energy that was emitted during the interval of after in the energy range. In the case of GRB 910402 only of the GRB energy was released during the observed afterglow ( after ).

Fig. 3. The light curve of GRB 910402 in energy range in logarithmic coordinates. The instrumental background is subtracted. corresponds to the best fit estimate of the beginning of the afterglow emission. The dashed curve corresponds to the law for which the afterglow begins at moment . Note that power line is curved in Log-Log space because the beginning of coordinates in this figure () differs from moment .

Fig. 4. The evolution of hardness of GRB 910402 (The ratio of the count rates in to energy ranges). The values of effective photon index on the right axis corresponds to the hardness values (left axis) for the power law spectra. corresponds to the moment of the abrupt softening of the emission from GRB source.

Fig. 5. The same as in the Fig. 3 but for the GRB 920723

Fig. 6. The same as in the Fig. 4 but for the GRB 920723

© European Southern Observatory (ESO) 2000

Online publication: June 8, 2000
helpdesk.link@springer.de