Speaker
Description
The Compton Spectrometer and Imager (COSI), scheduled for launch in 2027, will improve sensitivity for MeV gamma-ray line observations by an order of magnitude, opening new windows to study cosmic nucleosynthesis. Primary targets include the 511 keV emission from positron annihilation, $^{26}$Al from stellar nucleosynthesis, and $^{44}$Ti from supernova remnants. With its wide field of view, COSI will provide all-sky images for these lines with the best sensitivity ever. However, image reconstruction in this energy band presents unique challenges: an event of Compton telescopes constrains the gamma-ray direction to a circle in the sky rather than a point, necessitating a statistical approach to recover the gamma-ray source distribution. Also, the complex background environment requires careful treatment for reliable source detection.
We present an advanced image reconstruction framework for MeV gamma-ray observations. Our approach introduces a modified Richardson-Lucy algorithm with Bayesian priors optimized to address these challenges. The framework flexibly incorporates prior distributions, such as the sparseness and smoothness of an image, while simultaneously fitting the background components. We evaluate the method using simulated three-month COSI observations of key nuclear lines ($^{44}$Ti at 1.157 MeV, $^{26}$Al at 1.809 MeV, and positron annihilation at 0.511 MeV). Our results show successful suppression of artifacts in point source reconstructions while preserving extended emission features, demonstrating improved image quality and sensitivity compared to conventional techniques. This work represents an important step toward advancing our understanding of nucleosynthesis, positron annihilation, and other high-energy phenomena with future MeV gamma-ray line observations.
| Collaboration(s) | The COSI collaboration |
|---|
