All Figures

$\begin{figure} \par\resizebox{8.3cm}{!}{\includegraphics{0511_01a.eps}}\par\vspace*{2mm} \resizebox{8.3cm}{!}{\includegraphics{0511_01b.ps}} \end{figure}$ Figure 1: 1-s background subtracted light curves of the two flares in the 40-700 keV energy band. Top: 2001 April 18 event; bottom: 1998 August 27 event. The 5.2-s modulation is apparent in both cases. Vertical dotted lines mark the 128-s intervals during which 225-channel spectra were accumulated on board.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8cm,clip]{0511_02.ps} \end{figure}$ Figure 2: Background subtracted light curve of IF01 compared with the corresponding first 40 s of GF98 in the 40-700 keV energy band. Top: IF01; bottom: GF98. Vertical dotted lines mark the 5.2-s intervals and are synchronized with the IF01 pulsations.
Open with DEXTER

In the text

$\begin{figure} \par\resizebox{8cm}{!}{\includegraphics{0511_03.eps}} \end{figure}$ Figure 3: Light curves of the first $\sim$ 40 s of both IF01 ( black) and GF98 ( red) superposed to each other assuming that the two steepest rises correspond to the onset of the flares.
Open with DEXTER

In the text

$\begin{figure} \par\resizebox{8cm}{!}{\includegraphics{0511_04.ps}} \end{figure}$ Figure 4: Light curves of the first $\sim$ 40 s of both IF01 ( black) and GF98 ( red) superposed to each other, with the first interpulses aligned. The slices' boundaries have been chosen to mark the first minima.
Open with DEXTER

In the text

$\begin{figure} \par {\includegraphics[width=8.8cm,clip]{0511_05.eps} }\end{figure}$ Figure 5: Precursor in the 40-700 keV light curve of GF98 at $t\sim -0.45$ s, followed by a slow increase in intensity just before the initial pulse. We added the counts of all GRBM units to improve the statistical quality.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=16.3cm,clip]{0511_06.eps} \end{figure}$ Figure 6: Pulse shape evolution of both flares (IF01: black, GF98: red) during the first nine slices. The time offset between the two events was chosen in order to align the first interpulses (see also Fig. 4). Horizontal scale: phase of rotation cycle; vertical scale: counts per 125 ms.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.2cm,clip]{0511_07.eps} \end{figure}$ Figure 7: IF01 light curve in the >100 keV energy band. The fact that the pulse preceding the main rise is also apparent in this band, seems to suggest that it is unlikely to be a precursor.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[height=7.0cm,width=5.0cm,angle=270,clip]{0511_08.ps} \end{figure}$ Figure 8: Measured PSD of IF01 using as time binning 7.8125 ms. Solid line: PSD of the best-fit exponential function used to describe the mean light curve of IF01, excluding the precursor (see text). Dashed line: Poissonian level corrected for the dead time.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[height=7.0cm,width=5.0cm,angle=270,clip]{0511_09.ps} \end{figure}$ Figure 9: 20 Hz-1 kHz PSD of IF01 measured in the time interval [+0.7,+10.7] s after subtraction of the non-stationary Poissonian noise dead-time corrected.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[height=7.0cm,width=5.0cm,angle=270,clip]{0511_10.ps} \end{figure}$ Figure 10: PSD of the first 38 s of GF98 with time origin 1 s after the onset, in order to exclude the spike. The time binning used is 7.8125 ms. Solid line: PSD of the best-fit exponential function used to describe the mean light curve. Dashed line: Poissonian level corrected for the dead time.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[height=7.0cm,width=5.0cm,angle=270,clip]{0511_11.ps} \end{figure}$ Figure 11: PSD of GF98 in the time interval [+1, +9] s. Continuum line: PSD of the best-fit exponential function used to describe the first 10 s of the mean light curve. Dashed line: Poissonian level corrected for dead time. Inset: zoomed high-frequency PSD.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[height=7.0cm,width=5.0cm,angle=270,clip]{0511_12.ps} \end{figure}$ Figure 12: PSD of the residual noise of GF98 in the time interval [1-9] s. Solid line: best-fit power-law (see text).
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[height=7.0cm,width=5.0cm,angle=270,clip]{0511_13.ps} \end{figure}$ Figure 13: PSD of GF98 for the interval [+1-+9] s averaged over the four GRBM units. Solid line: best-fit power law (see text).
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=5.0cm,angle=270,clip]{0511_14a.ps}\par... ...ace*{2mm} \includegraphics[width=5.0cm,angle=270,clip]{0511_14b.ps} \end{figure}$ Figure 14: Count rate spectra of IF01 and their fit with a BB + BKNPL during the interval A, and with a BB model during the interval B. The residuals to the models are shown as well.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=5.0cm,angle=270,clip]{0511_15a.ps}\par... ...ace*{2mm} \includegraphics[width=5.0cm,angle=270,clip]{0511_15c.ps} \end{figure}$ Figure 15: Count rate spectra of GF98 and their fit with a BB + BKNPL during the interval A and B, and with a BB model during the interval C. The residuals to the model are shown as well.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=5.0cm,angle=270,clip]{0511_16a.ps}\par... ...ace*{2mm} \includegraphics[width=5.0cm,angle=270,clip]{0511_16b.ps} \end{figure}$ Figure 16: EF(E) spectra of IF01. Top panel: A spectrum along with the best fit with a BB+BKNPL+PL model. Bottom panel: B spectrum along with the fit with a BB. Also the single model components and the residuals (in units of $\sigma$ ) to the best-fit models are shown.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=5.0cm,angle=270,clip]{0511_17a.ps}\par... ...ace*{2mm} \includegraphics[width=5.0cm,angle=270,clip]{0511_17c.ps} \end{figure}$ Figure 17: EF(E) average spectra of GF98. Top panel: A spectrum with superposed the best-fit model BB+BKNPL+PL. Middle panel: B spectrum with superposed the best-fit model BB+BKNPL. Bottom panel: C spectrum with superposed the best-fit model BB+PL. Also the model components and the residuals to the best-fit models are shown.
Open with DEXTER

In the text

$\begin{figure} \par {\hspace*{2mm}\includegraphics[width=5.0cm,angle=270,clip]{0... ...ace*{2mm} \includegraphics[width=5.0cm,angle=270,clip]{0511_18b.ps} \end{figure}$ Figure 18: Time behaviour of the equivalent kT of an OTTB model for the entire duration of IF01 and for the early part of GF98. Error bars are 90% CL. Top panel: GF98; bottom panel: IF01.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[height=7.5cm,angle=270,clip]{0511_19.ps} \end{figure}$ Figure 19: Light curve of IF01; the dashed line is the best fit with a fireball model obtained for the time interval [7.46875-38.5] s, rebinned at 5.171875 s; the solid line shows the best fit applied to the interval [3-38] s, whose parameters have been constrained to vary within 1 $\sigma$ around the best fit of the tail.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.5cm,clip]{0511_20.eps} \end{figure}$ Figure 20: 40-700 keV time profiles of the last isolated pulse of IF01 ( blue) and of the August 29, 1998 burst ( red).
Open with DEXTER

In the text

$\begin{figure} \par\resizebox{8.3cm}{!}{\includegraphics{0511_01a.eps}}\par\vspace*{2mm} \resizebox{8.3cm}{!}{\includegraphics{0511_01b.ps}} \end{figure}$	Figure 1: 1-s background subtracted light curves of the two flares in the 40-700 keV energy band. Top: 2001 April 18 event; bottom: 1998 August 27 event. The 5.2-s modulation is apparent in both cases. Vertical dotted lines mark the 128-s intervals during which 225-channel spectra were accumulated on board.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=8cm,clip]{0511_02.ps} \end{figure}$	Figure 2: Background subtracted light curve of IF01 compared with the corresponding first 40 s of GF98 in the 40-700 keV energy band. Top: IF01; bottom: GF98. Vertical dotted lines mark the 5.2-s intervals and are synchronized with the IF01 pulsations.
Open with DEXTER

$\begin{figure} \par\resizebox{8cm}{!}{\includegraphics{0511_03.eps}} \end{figure}$	Figure 3: Light curves of the first $\sim$ 40 s of both IF01 ( black) and GF98 ( red) superposed to each other assuming that the two steepest rises correspond to the onset of the flares.
Open with DEXTER

$\begin{figure} \par\resizebox{8cm}{!}{\includegraphics{0511_04.ps}} \end{figure}$	Figure 4: Light curves of the first $\sim$ 40 s of both IF01 ( black) and GF98 ( red) superposed to each other, with the first interpulses aligned. The slices' boundaries have been chosen to mark the first minima.
Open with DEXTER

$\begin{figure} \par {\includegraphics[width=8.8cm,clip]{0511_05.eps} }\end{figure}$	Figure 5: Precursor in the 40-700 keV light curve of GF98 at $t\sim -0.45$ s, followed by a slow increase in intensity just before the initial pulse. We added the counts of all GRBM units to improve the statistical quality.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=16.3cm,clip]{0511_06.eps} \end{figure}$	Figure 6: Pulse shape evolution of both flares (IF01: black, GF98: red) during the first nine slices. The time offset between the two events was chosen in order to align the first interpulses (see also Fig. 4). Horizontal scale: phase of rotation cycle; vertical scale: counts per 125 ms.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.2cm,clip]{0511_07.eps} \end{figure}$	Figure 7: IF01 light curve in the >100 keV energy band. The fact that the pulse preceding the main rise is also apparent in this band, seems to suggest that it is unlikely to be a precursor.
Open with DEXTER

$\begin{figure} \par\includegraphics[height=7.0cm,width=5.0cm,angle=270,clip]{0511_08.ps} \end{figure}$	Figure 8: Measured PSD of IF01 using as time binning 7.8125 ms. Solid line: PSD of the best-fit exponential function used to describe the mean light curve of IF01, excluding the precursor (see text). Dashed line: Poissonian level corrected for the dead time.
Open with DEXTER

$\begin{figure} \par\includegraphics[height=7.0cm,width=5.0cm,angle=270,clip]{0511_09.ps} \end{figure}$	Figure 9: 20 Hz-1 kHz PSD of IF01 measured in the time interval [+0.7,+10.7] s after subtraction of the non-stationary Poissonian noise dead-time corrected.
Open with DEXTER

$\begin{figure} \par\includegraphics[height=7.0cm,width=5.0cm,angle=270,clip]{0511_10.ps} \end{figure}$	Figure 10: PSD of the first 38 s of GF98 with time origin 1 s after the onset, in order to exclude the spike. The time binning used is 7.8125 ms. Solid line: PSD of the best-fit exponential function used to describe the mean light curve. Dashed line: Poissonian level corrected for the dead time.
Open with DEXTER

$\begin{figure} \par\includegraphics[height=7.0cm,width=5.0cm,angle=270,clip]{0511_11.ps} \end{figure}$	Figure 11: PSD of GF98 in the time interval [+1, +9] s. Continuum line: PSD of the best-fit exponential function used to describe the first 10 s of the mean light curve. Dashed line: Poissonian level corrected for dead time. Inset: zoomed high-frequency PSD.
Open with DEXTER

$\begin{figure} \par\includegraphics[height=7.0cm,width=5.0cm,angle=270,clip]{0511_12.ps} \end{figure}$	Figure 12: PSD of the residual noise of GF98 in the time interval [1-9] s. Solid line: best-fit power-law (see text).
Open with DEXTER

$\begin{figure} \par\includegraphics[height=7.0cm,width=5.0cm,angle=270,clip]{0511_13.ps} \end{figure}$	Figure 13: PSD of GF98 for the interval [+1-+9] s averaged over the four GRBM units. Solid line: best-fit power law (see text).
Open with DEXTER

$\begin{figure} \par\includegraphics[width=5.0cm,angle=270,clip]{0511_14a.ps}\par... ...ace*{2mm} \includegraphics[width=5.0cm,angle=270,clip]{0511_14b.ps} \end{figure}$	Figure 14: Count rate spectra of IF01 and their fit with a BB + BKNPL during the interval A, and with a BB model during the interval B. The residuals to the models are shown as well.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=5.0cm,angle=270,clip]{0511_15a.ps}\par... ...ace*{2mm} \includegraphics[width=5.0cm,angle=270,clip]{0511_15c.ps} \end{figure}$	Figure 15: Count rate spectra of GF98 and their fit with a BB + BKNPL during the interval A and B, and with a BB model during the interval C. The residuals to the model are shown as well.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=5.0cm,angle=270,clip]{0511_16a.ps}\par... ...ace*{2mm} \includegraphics[width=5.0cm,angle=270,clip]{0511_16b.ps} \end{figure}$	Figure 16: EF(E) spectra of IF01. Top panel: A spectrum along with the best fit with a BB+BKNPL+PL model. Bottom panel: B spectrum along with the fit with a BB. Also the single model components and the residuals (in units of $\sigma$ ) to the best-fit models are shown.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=5.0cm,angle=270,clip]{0511_17a.ps}\par... ...ace*{2mm} \includegraphics[width=5.0cm,angle=270,clip]{0511_17c.ps} \end{figure}$	Figure 17: EF(E) average spectra of GF98. Top panel: A spectrum with superposed the best-fit model BB+BKNPL+PL. Middle panel: B spectrum with superposed the best-fit model BB+BKNPL. Bottom panel: C spectrum with superposed the best-fit model BB+PL. Also the model components and the residuals to the best-fit models are shown.
Open with DEXTER

$\begin{figure} \par {\hspace{2mm}\includegraphics[width=5.0cm,angle=270,clip]{0... ...ace{2mm} \includegraphics[width=5.0cm,angle=270,clip]{0511_18b.ps} \end{figure}$	Figure 18: Time behaviour of the equivalent kT of an OTTB model for the entire duration of IF01 and for the early part of GF98. Error bars are 90% CL. Top panel: GF98; bottom panel: IF01.
Open with DEXTER

$\begin{figure} \par\includegraphics[height=7.5cm,angle=270,clip]{0511_19.ps} \end{figure}$	Figure 19: Light curve of IF01; the dashed line is the best fit with a fireball model obtained for the time interval [7.46875-38.5] s, rebinned at 5.171875 s; the solid line shows the best fit applied to the interval [3-38] s, whose parameters have been constrained to vary within 1 $\sigma$ around the best fit of the tail.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.5cm,clip]{0511_20.eps} \end{figure}$	Figure 20: 40-700 keV time profiles of the last isolated pulse of IF01 ( blue) and of the August 29, 1998 burst ( red).
Open with DEXTER