Detection of the 511 keV Galactic positron annihilation line with COSI

TL;DR

This paper reports the detection and analysis of the 511 keV positron annihilation line from the Galactic center using the COSI balloon-borne telescope, revealing a broader emission distribution than previously observed.

Contribution

First detection of the 511 keV line with COSI, employing a novel background separation technique, and providing new insights into the spatial distribution of Galactic positron annihilation.

Findings

7.2σ detection of the 511 keV line

Emission distribution is broader than previously reported

No single point source accounts for the emission

Abstract

The signature of positron annihilation, namely the 511 keV $\gamma$-ray line, was first detected coming from the direction of the Galactic center in the 1970's, but the source of Galactic positrons still remains a puzzle. The measured flux of the annihilation corresponds to an intense steady source of positron production, with an annihilation rate on the order of $\sim10^{43}$~e$^{+}$/s. The 511 keV emission is the strongest persistent Galactic $\gamma$-ray line signal and it shows a concentration towards the Galactic center region. An additional low-surface brightness component is aligned with the Galactic disk; however, the morphology of the latter is not well constrained. The Compton Spectrometer and Imager (COSI) is a balloon-borne soft $\gamma$-ray (0.2--5 MeV) telescope designed to perform wide-field imaging and high-resolution spectroscopy. One of its major goals is to further our understanding of Galactic positrons. COSI had a 46-day balloon flight in May--July 2016 from Wanaka, New Zealand, and here we report on the detection and spectral and spatial analyses of the 511 keV emission from those observations. To isolate the Galactic positron annihilation emission from instrumental background, we have developed a technique to separate celestial signals utilizing the COMPTEL Data Space. With this method, we find a 7.2$\sigma$ detection of the 511 keV line. We find that the spatial distribution is not consistent with a single point source, and it appears to be broader than what has been previously reported.