The emission processes active in the highly relativistic jets of gamma-ray bursts (GRBs) remain unknown. In this paper we propose a new measure to describe spectra: the width of the $EF_E$ spectrum, a quantity dependent only on finding a good fit to the data. We apply this to the full sample of GRBs observed by Fermi/GBM and CGRO/BATSE. The results from the two instruments are fully consistent. We find that the median widths of spectra from long and short GRBs are significantly different (chance probability $<10^{-6}$). The width does not correlate with either duration or hardness, and this is thus a new, independent distinction between the two classes. Comparing the measured spectra with widths of spectra from fundamental emission processes -- synchrotron and blackbody radiation -- the results indicate that a large fraction of GRB spectra are too narrow to be explained by synchrotron radiation from a distribution of electron energies: for example, 78% of long GRBs and 85% of short GRBs are incompatible with the minimum width of standard slow cooling synchrotron emission from a Maxwellian distribution of electrons, with fast cooling spectra predicting even wider spectra. Photospheric emission can explain the spectra if mechanisms are invoked to give a spectrum much broader than a blackbody.

A sample of 197 long BATSE GRBs is studied statistically. In the sample 11 variables, describing for any burst the time behavior of the spectra and other quantities, are collected. The application of the factor analysis on this sample shows that five factors describe the sample satisfactorily. Both the pseudo-redshifts coming from the variability and the Amati-relation in its original form are disfavored.

We study statistically 197 long gamma-ray bursts, detected and measured in detail by the BATSE instrument of the Compton Gamma-Ray Observatory. In the sample 10 variables, describing for any burst the time behavior of the spectra and other quantities, are collected. The factor analysis method is used to find the latent random variables describing the temporal and spectral properties of GRBs. The application of this particular method to this sample indicates that five factors and the $\REpk$ spectral variable (the ratio of peak energies in the spectrum) describe the sample satisfactorily. Both the pseudo-redshifts inferred from the variability, and the Amati-relation in its original form, are disfavored.

Aims. We investigate observations of the X-ray binary Cygnus X-1 with unusually high hardness and low flux. In particular, we study the characteristic frequencies seen in the PDS and the hardness-flux correlation within and between these observations. Methods. We analyse observations of Cyg X-1 during periods when the source reaches its highest hardness levels (> 1 for the 9-20 keV over 2-4 keV RXTE/PCA count ratios, corresponding to Gamma < 1.6). Using the relativistic precession model to interpret the PDS we estimate a value for the inner radius of the accretion disc. We also study the hardness-flux correlation. Results. In the selected observations, the characteristic frequencies seen in the power spectrum are shifted to the lowest end of their frequency range. Within a single observation, the hardness-flux correlation is very weak, contrary to the negative correlation normally observed in the hard state. We suggest that this could be interpreted as the inner disc boundary being at large radii (> 50 Rg), thereby requiring more time to adjust to a changing accretion rate than allowed by a single RXTE observation, and compare our findings to estimates of the viscous time scale responsible for small scale variability in the system.

Principal component analysis is a statistical method, which lowers the number of important variables in a data set. The use of this method for the bursts' spectra and afterglows is discussed in this paper. The analysis indicates that three principal components are enough among the eight ones to describe the variablity of the data. The correlation between spectral index alpha and the redshift suggests that the thermal emission component becomes more dominant at larger redshifts.

The origin of dark bursts - i.e. that have no observed afterglows in X-ray, optical/NIR and radio ranges - is unclear yet. Different possibilities - instrumental biases, very high redshifts, extinction in the host galaxies - are discussed and shown to be important. On the other hand, the dark bursts should not form a new subgroup of long gamma-ray bursts themselves.

Based on the analysis of a small sample of BATSE and Konus gamma-ray bursts (GRBs) with know redshift it has been reported that the width of the autocorrelation function (ACF) shows a remarkable bimodal distribution in the rest-frame of the source. However, the origin of these two well-separated ACF classes remains unexplained.We complement previous ACF analysis studying the corresponding power density spectra (PDS). With the addition of Beppo-SAX data and taken advantage of its broad-band capability, we not only increase the burst sample but we extend the analysis to X-ray energies. The rest-frame PDS analysis at gamma-ray energies shows that the two ACF classes are not simply characterised by a different low frequency cut-off, but they have a distinct variability as a whole in the studied frequency range. Both classes exhibit average PDS with power-law behaviour at high frequencies (f' > 0.1 Hz) but significantly different slopes, with index values close to those of Brownian (-2) and Kolmogorov (-5/3) spectra for the narrow and broad classes respectively. The latter spectrum presents an additional PDS component, a low-frequency noise excess with a sharp cut-off. At X-ray energies we find the power-law index unchanged for the broad class, but a significantly steeper slope in the narrow case (~ -3). We interpret this as an indication that the broad class bursts have weaker spectral evolution than the narrow ones, as suggested also by our analysis of the ACF energy dependence. The low and high frequency PDS components may then arise from two radiating regions involving different emission mechanisms.

The statistical properties of a complete, flux limited sample of 197 long gamma--ray bursts (GRBs) detected by BATSE are studied. In order to bring forth their main characteristics, care was taken to define a representative set of ten parameters. A multivariate analysis gives that ~70% of the total variation in parameter values is driven by only three principal components. The variation of the temporal parameters is clearly distinct from that of the spectral ones. A close correlation is found between the half-width of the autocorrelation function (tau) and the emission time (Tem); most importantly, this correlation is self-similar in the sense that the mean values and dispersions of both tau and Tem scale with the duration of the burst (T90). It is shown that the Amati-relation can be derived from the sample and that the scatter around this relation is correlated with the value of tau. Hence, tau has a role similar to that of the break in the afterglow light curve (Tb) in the Ghirlanda-relation. In the standard GRB-scenario, the close relation between a global parameter (Tb) and a local one (tau) indicates that some of the jet-properties do not vary much for different lines of sight. Finally, it is argued that the basic temporal and spectral properties are associated with individual pulses, while the overall properties of a burst is determined mainly by the number of pulses.

Building on results from previous studies of Cygnus~X-1, we analyze Rossi X-ray Timing Explorer (RXTE) data taken when the source was in the soft and transitional spectral states. We look at the power spectrum in the 0.01 -- 50 Hz range, using a model consisting of a cut-off power-law and two Lorentzian components. We are able to constrain the relation between the characteristic frequencies of the Lorentzian components, and show that it is consistent with a power-law relation having the same index (1.2) as previously reported for the hard state, but shifted by a factor ~2. Furthermore, it is shown that the change in the frequency relation seen during the transitions can be explained by invoking a shift of one Lorentzian component to a higher harmonic, and we explore the possible support for this interpretation in the other component parameters. With the improved soft state results we study the evolution of the fractional variance for each temporal component. This approach indicates that the two Lorentzian components are connected to each other, and unrelated to the power-law component in the power spectrum, pointing to at least two separate emission components.

Analyzing the archival data from the Rossi X-ray Timing Explorer (RXTE), we study the power density spectra (PDS) of Cygnus X-1 from 1996 to 2003 in the frequency range of 0.01 - 25 Hz. With our model we are able to track the evolution of the Lorentzian components through all spectral states of the source. We confirm the relation between characteristic frequencies seen both in black hole candidate and neutron star sources, and show the changes in this relation during the transitional and soft states of the source. The connection between the Lorentzian components is investigated by analyzing similarities and differences in their behavior. We find that the spectral state of the source can be uniquely determined from the parameters of the these components. The parameter correlations can all be described by continuous functions, which differ between components. We discuss our results in the context of relativistic precession model for the accretion disk, and show a remarkable agreement between the model prediction and the data in the hard state. We estimate a value for the specific angular momentum of a=+0.49 (-0.57) in the case of prograde retrograde) rotation and an estimate for the inner radius of 22 to 50 (25 to 55) gravitational radii.

Autocorrelation functions (ACFs) are studied for a sample of 16 long gamma-ray bursts (GRBs) with known redshift z, that were observed by the BATSE and Konus experiments. When corrected for cosmic time dilation, the ACF shows a bimodal distribution. A narrow width class (11 bursts) has at half-maximum a mean width of 1.6 s with a relative dispersion of 32%, while a broad width class (5 bursts) has a width of 7.5 s with a 4% dispersion. The separation between the two mean values is highly significant (> 7 standard deviations). This temporal property could be used on the large existing database of GRBs with unknown redshift. The broad width set shows a very good linear correlation between width at half-maximum and (1+z), with a correlation coefficient R=0.995 and a probability of chance alignment <0.0004. The potential application of this correlation to cosmology studies is discussed, using it in combination with recently proposed luminosity indicators.

We compare the energy and count fluxes obtained by integrating over the finite bandwidth of BATSE with a measure proportional to the bolometric energy flux, the phi-measure, introduced by Borgonovo & Ryde. We do this on a sample of 74 bright, long, and smooth pulses from 55 GRBs. The correction factors show a fairly constant behavior over the whole sample, when the signal-to-noise-ratio is high enough. We present the averaged spectral bolometric correction for the sample, which can be used to correct flux data.

We study the hardness-intensity correlation (HIC) in gamma-ray bursts (GRBs). In particular, we analyze the decay phase of pulse structures in their light curves. The study comprises a sample of 82 long pulses selected from 66 long bursts observed by BATSE on the Compton Gamma-Ray Observatory. We find that at least 57% of these pulses have HICs that can be well described by a power law. The distribution of the power law indices, obtained by modeling the HIC of pulses from different bursts, is broad with a mean of 1.9 and a standard deviation of 0.7. We also compare indices among pulses from the same bursts and find that their distribution is significantly narrower. The probability of a random coincidence is shown to be very small. In most cases, the indices are equal to within the uncertainties. This is particularly relevant when comparing the external versus the internal shock models. In our analysis, we also use a new method for studying the HIC, in which the intensity is represented by the peak value of the E F_E spectrum. This new method gives stronger correlations and is useful in the study of various aspects of the HIC. In particular, it produces a better agreement between indices of different pulses within the same burst. Also, we find that some pulses exhibit a "track jump" in their HICs, in which the correlation jumps between two power laws with the same index. We discuss the possibility that the "track jump" is caused by strongly overlapping pulses. Based on our findings, the constancy of the index is proposed to be used as a tool for pulse identification in overlapping pulses.