XMM-Newton Trainee Project 2004

Maria Pilar Esquej Alonso

I was born in Zaragoza, Spain. After studying three years of Physics at the Universidad de Zaragoza, I moved to Madrid in order to finish my degree in the specialization field of Astrophysics at the Universidad Complutense de Madrid, where I graduated in September of 2004.

This document describes a brief overview of the project that I performed during six months, forming part of the pre-graduate program for Spanish students at the XMM-Newton Science Operations Centre, ESAC.

On board the ESA X-ray observatory XMM-Newton, three different CCD-cameras comprising the European Photon Imaging Camera (EPIC) are specially developed for X-ray imaging and spectroscopy to observe the hot universe in an extremely sensitive way. The accuracy of the energy cross-calibration of the EPIC-pn camera in different operating modes was determined, using observations of the line-rich supernova remnant (SNR) Cas-A.

For the Extended Full Frame (eFF) mode a special energy calibration was performed, which showed an overcorrection of the energy with respect to the Full Frame (FF) mode for the internal calibration source at aluminium and manganese energies. A dedicated calibration observation of Cas-A performed in revolution 836 was analysed, and spectrum of the target was extracted using the XMM-Newton Science Analysis System (SAS), in order to correct the energy differences in a wide energy X-ray range. Using the standard X-ray spectral fitting tool XSPEC line positions of the prominent Cas-A lines were determined. Ratios of line positions between the two modes were used to derive a smooth Boltzmann function of the energy (Fig. 1) that was implemented in the CAL (version 3.172.5), and corresponding parameters were added to the EPN_CTI_0014.CCF in the extension eFF_Gain. After reprocessing the data with the updated SAS (6.1.0), accuracy in line energy determination in eFF mode was supposed to be better than 0.1 %. Using the same procedure with observations of Cas-A performed in different revolutions and pn modes (Full Frame, Large Window, Small Window, Timing), line positions were determined and ratios respect to the FF mode were calculated. This showed there was an imprecision in line energy determination of 0.1%, except for the Small Window mode, whose line determination of 0.4% was due to a gain/CTI undercorrection. All determined ratios of line positions are showed in Fig. 2.




Figure 1. Gain tuning function to adjust eFF to FF gain



Figure 2. Ratio of line positions versus energy between different modes.