Title: Improving INTEGRAL/SPI data analysis of GRBs

Abstract: INTEGRAL/SPI is a coded mask instrument observing since 2002 in the keV to MeV energy range, which covers the peak of the $\nu F\nu$ spectrum of most Gamma-Ray Bursts (GRBs). Since its launch in 2008, Fermi/GBM has been the primary instrument for analyzing GRBs in the energy range between $\approx$ 10 keV to $\approx$ 10 MeV. Herein, we show that SPI, covering a similar energy range, can give equivalently constraining results for some parameters if we use an advanced analysis method. Also, combining the data of both instruments reduces the allowed parameter space in spectral fits. The main advantage of SPI as compared to GBM is the energy resolution of $\approx$ 0.2\% at 1.3 MeV compared to $\approx$ 10\% for GBM. Therefore, SPI is an ideal instrument to precisely measure the curvature of the spectrum. This is important, as it has been shown in recent years that physical models rather than heuristic functions should be fit to GRB data to obtain better insights into their still unknown emission mechanism, and the curvature of the peak is unique to the different physical models. To fit physical models to SPI GRB data and get the maximal amount of information from the data, we developed a new open source analysis software {\tt PySPI}. We apply these new techniques to GRB 120711A in order to validate and showcase {\tt PySPI}'s capabilities. We show that {\tt PySPI} improves the analysis of SPI GRB data compared to the {\tt OSA} analysis. In addition, we demonstrate that the GBM and the SPI data of this GRB can be fitted well with a physical synchrotron model. This evinces that SPI can play an important role in GRB spectral model fitting.