The radiative mechanism responsible for the prompt gamma-ray burst (GRB) emission remains elusive. For the last decade, optically thin synchrotron emission from shocks internal to the GRB jet appeared to be the most plausible explanation. However, the synchrotron interpretation is incompatible with a significant fraction of GRB observations, highlighting the need for new ideas. In this thesis, it is shown that the narrow, dominating component of the prompt emission from the bright GRB090902B is initially consistent only with emission released at the optically thick jet photosphere. However, this emission component then broadens in time into a more typical GRB spectrum, which calls for an explanation. In this thesis, a previously unconsidered way of broadening the spectrum of photospheric emission, based on considerations of the lateral jet structure, is presented and explored. Expressions for the spectral features, as well as polarization properties, of the photospheric emission observed from structured, relativistic jets are derived analytically under simplifying assumptions on the radiative transfer close to the photosphere. The full, polarized radiative transfer is solved through Monte Carlo simulations, using a code which has been constructed for this unique purpose. It is shown that the typical observed GRB spectrum can be obtained from the photosphere, without the need for additional, commonly assumed, physical processes (e.g. energy dissipation, particle acceleration, or additional radiative processes). Furthermore, contrary to common expectations, it is found that the observed photospheric emission can be highly linearly polarized (up to $\sim 40 \, \%$). In particular, it is shown that a shift of $\pi/2$ of the angle of polarization is the only shift allowed by the proposed model, consistent with the only measurement preformed to date. A number of ways to test the theory is proposed, mainly involving simultaneous spectral and polarization measurements. The simplest measurement, which tests not only the proposed theory but also common assumptions on the jet structure, involves only two consecutive measurements of the angle of polarization during the prompt emission. / <p>QC 20131204</p>

Gamma-ray bursts (GRBs) are the most energetic electromagnetic events in the universe, but there are still unanswered questions about them, like the underlying radiation mechanisms that cause the different parts of their light curves. Given that Wolf-Rayet (WR) stars with circumburst rings could be the progenitor of GRBs with late flares \cite{complex}, the purpose of this thesis was to determine if the precursor and the main emission of GRBs with late time flares might originate due to different radiation mechanisms, and thereby if WR stars could be the progenitors. 271 of the longest GRBs with flux above 10 photons/cm$^2$/s were studied and all GRBs with a precursor and a defined quiescent period were chosen for further studies. The chosen 39 GRBs were divided into different categories depending on the appearance of their light curves. A gold sample with $R_{P, max}/R_{D, max} &lt; 0.4$ and $T_Q/T_{tot} &gt; 0.5$, for the maximum count rate of the precursor $R_{P, max}$, the dominant emission $R_{D, max}$, and for the normalised quiescent period $T_Q/T_{tot}$ was concluded to have similar characteristics and to fit what would be expected if WR stars would be the progenitors. This group of GRBs all have a short and less bright precursor, a long quiescent period and a main emission which is brighter and longer than the precursor. The distributions of the photon index $\alpha$ for the precursor and the dominant emission for the gold sample indicate that the precursor is due to photospheric emission and the dominant emission is due to synchrotron emission. This is consistent with the interpretation that the precursor is a result of the jet interacting with the photosphere and the dominant emission is a result of interactions with the circumburst ring of a star like the WR stars. The next step in this investigation would be to study GRBs with more than one precursor that otherwise fit the description of the gold sample, to determine if these fit into the gold sample as well. / Gammablixtrar (GRB) är de mest kraftfulla elektromagnetiska eventen i universum men det finns fortfarande obesvarade frågor om dem, som de underliggande strålningsmekanismerna som orsakar de olika delarna av deras ljuskurvor. För en del av alla GRBs tar det upp till eller mer än 100 sekunder från utlösningstiden till det att en topp ses i ljuskurvan. För dessa finns då ofta en liten svag topp, som följs av en lång lugn period och sedan den dominant, starkare utstrålningen. GRBs tros kunna härstamma från Wolf-Rayet-stjärnor (WR-stjärnor), som är massiva, döende stjärnor som kan vara omringade av bubblor, nebulosor, och ringar. Om GRBs härstammar från dessa förväntas den första mindre toppen och den andra större toppen uppkomma på grund av olika strålningsprocesser. Syftet med detta projekt var därför att undersöka huruvida dessa toppar uppkommer på grund av olika processer eller ej. Sammanfattningsvis så hittades en distinkt och homogen grupp av GRBs med likande egenskaper. Resultaten påvisar att den första svaga toppen är fotosfärisk strålning, så att den uppkommer på grund av att jetstrålen från GRBn interagerar med fotosfären. Därtill tyder resultaten på att den dominanta starkare toppen är synkrotronstrålning, som kan uppkomma när jetstrålen interagerar med en ring runt en WR-stjärna. Nästa steg i detta projekt skulle vara att studera GRBs med fler än en mindre topp innan den dominant utstrålningen, för att se om dessa också har liknande egenskaper som de som hittades i den homogena gruppen.