The ejecta following the collision and merging of two neutron stars (kilonova) are currently considered promising sites for nucleosynthesis of r-process elements. Since the observed kilonova in 2017, GW2017817, astrophysicists have been working to analyze the collected electromagnetic spectra, trying to identify r-process elements. However, a lack of fundamental atomic data has been holding the efforts back. Motivated by spectral modelling of kilonovae out of equilibrium, this project aims to create a tool that uses line lists of spectroscopic accuracy from the Vienna Atomic Line Database (VALD) to generate energy level lists automatically for any ion. VALD in particular is used because it has wavelengths accurate enough for line identification purposes. The resulting level lists are compared to equivalent level lists from the database managed by the National Institute for Standards and Technology (NIST), in order to both ensure that the tool worked, and identify discrepancies between the databases. A number of problems with the VALD data were identified, mostly resulting in duplicate and missing energy levels. Finally, we also test the data in computations of kilonova expansion opacities in a complete solar r-process abundance mixture. Further work is needed to evaluate how damaging these problems are when modelling kilonovae, and when necessary remedy said problems. / Ejektat från en kollision och sammanslagning av två neutronstjärnor (kilonova) betraktas som lovande platser för nukleosyntes av r-processämnena. Sedan den observerade kilonovan år 2017, GW2017817, har astrofysiker försökt analysera de insamlade elektromagnetiska spektrumen för att försöka identifiera r-processämnen. Denna analys har dock hindrats på grund av en brist på fundamental atomisk data. Motiverat av spektralmodellering av kilonovor utanför ekvilibrium, syftar detta projekt på att utveckla ett verktyg för att utifrån spektroskopiskt noggranna linjelistor från Vienna Atomic Line Database (VALD) gererera listor över energinivåer automatiskt för godtycklig jon. VALD används på grund av att den har våglänger som är noggranna nog för linjeidentifiering. De resulterande nivålistorna jämförs med motsvarande nivålistor från databasen som sköts av National Institute for Standards and Technology (NIST), detta för att dels säkerställa att verktyget fungerade, dels identifiera skillnader mellan databaserna. Ett antal problem med VALD identifierades, vilka oftast resulterade i dubbletter eller avsaknad av energinivåer. Slutligen testades datan i beräkningar av kilonova-expansionsopaciteter i en komplett solär r-process-ämnesblandning. Vidare arbete krävs för att evaluera hur skadliga dessa problem är för modellering av kilonovor, och vid behov åtgärda problemen.

<p dir="ltr">Neutron star mergers generate powerful gravitational waves and various types of electromagnetic signals, including gamma-ray bursts (GRB), kilonovae, and their afterglows. Observing and modeling these signals help us understand the physical processes of the merger events. Radiation from mergers can also serve as probes to study nuclear physics and cosmology. In this report, I focus on two types of signals: the GRB afterglow and the kilonova. GRB afterglows are non-thermal radiation produced by the interaction of relativistic jets and circumburst material, where the jets are launched perpendicular to the merger plane. Kilonovae are the thermal radiation emitted from the hot materials ejected during the merger. Besides the modeling of these objects, I also investigate their application in multimessenger astrophysics, especially the constraint on the expansion rate of the Universe. </p><p dir="ltr">First, I developed a GRB afterglow model to account for the off-axis observation of a structured jet. Using a jet structure derived from a three-dimensional general relativistic magnetohydrodynamic simulation, we performed a joint analysis of the multimessenger data of the neutron star merger event GW170817, including the gravitational wave data and GRB afterglow data in the radio band. We have tightly constrained the observing angle of GW170817 and broken the degeneracy between the inclination angle and luminosity distance measured in gravitational waves. With a better constrained distance, we improved the standard siren measurement of the Hubble constant to $H_0 = 69.5\pm 4\ \mathrm{km\ s^{-1}\ Mpc^{-1}}$. The error bar has been reduced by a factor of 2. This work demonstrates that the modeling of off-axis GRB afterglow can significantly improve the standard siren method, provided that we have a reliable jet structure.</p><p dir="ltr">Second, I upgrade the GRB afterglow model in the first work, extending it to the late time where lateral spreading of the GRB jet becomes important. In this model, the ultra-relativistic blastwave is approximated by an infinitely thin two-dimensional surface. With this approximation, the hydrodynamic equations can be analytically integrated over the radius. Further assuming axial symmetry, the three-dimensional hydrodynamic simulation can be reduced to one dimension, which significantly increases the computational efficiency. We have compared our method to full numerical simulations and existing GRB afterglow modeling tools. The comparison shows good agreement and verifies our approach. Compared to these tools, our model has better flexibility and is applicable in a broader context. This method has been developed into a numerical code, \texttt{jetsimpy}, which we have provided to the community. It will serve as a powerful tool in the era of multimessenger astrophysics.</p><p dir="ltr">Finally, I investigate the possibility of long-lived massive neutron stars as neutron star merger remnants. A long-lived massive neutron star can inject a significant amount of energy into the merger ejecta, boosting the luminosity of kilonova by several orders of magnitude. However, this type of event has not yet been observed in optical sky surveys. We developed a boosted kilonova model with a detailed calculation of the photoionization process to better describe the efficiency of energy injection from spin down power to the ejecta. Our study found that boosted kilonovae, if commonly occurring, they should have already been observed given the accumulated time in sky surveys. As a result, the absence of detection implies that long-lived massive neutron stars as neutron star merger remnants are likely to be rare in the Universe.</p>