Improving the HESS-IU Camera Calibration

Abstract

Gamma-ray astronomy, the study of the most energetic form of electromagnetic radiation emitted from cosmic sources, has witnessed significant scientific and engineering accomplishments through the last decades. One of the leading gamma-ray observatories is the High Energy Stereoscopic System (H.E.S.S.), an array of five Imaging Atmospheric Cherenkov Telescopes located in Namibia. The telescopes measure the flux of gamma- rays using large mirrors which reflect the Cherenkov light produced by gamma-ray induced elecromagnetic showers on special cameras with pixels consisting of photomultiplier tubes (PMTs). It has been collecting data since 2002, allowing scientists to investigate gamma-rays with very high energies up to 30 TeV. The observatory’s operation is split into two phases: H.E.S.S. Phase I consists of four telescopes with a 12 m mirror diameter, while H.E.S.S. Phase II is characterised by the installation of a fifth 28 m telescope with a lower energy threshold in 2015. In order to decrease the threshold of the array altogether, as well as use newer technologies for more reliable camera performance, the cameras of the four H.E.S.S. I telescopes have been upgraded (HESS-IU). The most important component of the upgrade is the NECTAr readout chip, developed for the next generation observatory, CTA, and first tested on H.E.S.S.. It consists of an analogue memory cyclical buffer, from which values inside a specific readout window are swiftly read out whenever there is an active trigger signal. Since the reliability of the data depends on the proper function of the camera, regular calibration is a fundamental step in achieving the observatory’s scientific goals. In this thesis we propose some improvements in the calibration of the HESS-IU camera. The first one is a method to adjust the position of the readout window of each pixel towards lower energy signals, taking into consideration the intensity of the light pulses from the telescope’s Flat-Fielding unit. The second one is a technique for the calibration of the ratio factor between ADC counts and measured photoelectrons, in other words the gain of each PMT, under the hypothesis that there is a linear relationship between gain and PMT voltage. The two approaches use a global data fitting method, where samples of the intensity or charge of the same pixel are not considered independent and hence are fitted to a function in parallel instead of serially. The efficiency of the proposed schemes has been tested, giving improved time uniformity between telescopes and consistent results with the previously used algorithm, while speeding up and stabilising the relevant calibration processes. Another goal of this work is to show that it is possible to reconstruct incomplete data affected by a mis- calibrated readout window, as was the case after a hardware change. Using special calibration runs with light pulses of different intensities, we parametrise the charge of the whole signal of 16 ns, using the charge measured 9 ns around the peak of the signal. We then deduce reconstruction coefficients which can be used to recover the missing part of the data. The parametrisation is tested using observation data, showing that this approach has a great potential after accounting for real observation conditions (pedestals, night sky background, etc)