Pre-flight Background Estimates for COSI

The Compton Spectrometer and Imager (COSI) is a Compton telescope designed to survey the 0.2 - 5 MeV sky, consisting of a compact array of cross-strip germanium detectors. It is planned to be launched in 2027 into an equatorial low-Earth (530 km) orbit with a prime mission duration of 2 years. The observation of MeV gamma rays is dominated by background, mostly from extragalactic and atmospheric photon but also from the activation of the detector materials induced by cosmic-ray interactions. Thus, background simulation and identification are crucial for the data analysis. In this work we perform Monte Carlo simulations of the background for the first 3 months in orbit, and we extrapolate the results to 2 years in orbit, in order to determine the build-up of the activation due to long-lived isotopes. We determine the rates of events induced by the background that are reconstructed as Compton events in the simulated COSI data. We find that the extragalactic background photons dominate at low energies (<660 keV), while delayed activation from cosmic-ray primaries (proton/alpha) and albedo photons dominate at higher energies. As part of this work, a comparison at low latitude (<1 deg) between recent measurement of the SAA by the High-Energy Particle Detector (HEPD-01) on board the China Seismo-Electromagnetic Satellite (CSES-01) and the AP9/AE9 model has been made, showing an overestimation of the flux by a factor 9 by the model. The systematic uncertainties associated with these components are quantified. This work marks a major step forward in estimating and understanding the expected background rates for the COSI satellite mission.

💡 Research Summary

The Compton Spectrometer and Imager (COSI) is a MeV‑range Compton telescope designed to survey the sky from 0.2 to 5 MeV. It will be launched in 2027 into an equatorial low‑Earth orbit (530 km, inclination ≈0.25°) and will operate for a nominal two‑year mission. Because MeV γ‑ray observations are background‑limited, a detailed understanding of all background components is essential for sensitivity predictions and data analysis.

In this work the authors performed Monte‑Carlo simulations of the on‑orbit background for the first three months of the mission using the MEGAlib/Geant4 framework (Cosima). Input spectra for the extragalactic gamma‑ray background (EGB), atmospheric albedo photons and neutrons, and primary cosmic‑ray (CR) particles (protons, α‑particles, electrons, positrons) were generated with an updated version of the Cumani et al. (2019) code. Solar modulation for protons and α‑particles was applied via the HelMod tool for the expected March‑June 2027 solar minimum, while electron/positron fluxes were extrapolated from AMS‑02 measurements using a simple scaling ratio.

The authors accounted for the geomagnetic rigidity cutoff (Rcut) along the orbit by computing it every 15 s with the OTSO Python package based on SPENVIS orbital ephemerides. The average cutoff (12.6 GV) was used to fix the spectral shape, while the total flux was modulated in time according to the instantaneous Rcut. This approximation is justified because Rcut varies only between 10.4 and 14.5 GV for the chosen orbit.

A key part of the study is the treatment of activation. Two activation schemes are available in MEGAlib; the authors employed the “memory‑retain” method that tracks each isotope from creation to decay, thereby reproducing the build‑up of long‑lived isotopes during the flight. Only the first three months of orbit were simulated due to computational cost, and the results were extrapolated to the full two‑year mission to estimate the contribution from long‑lived isotopes such as ²⁴Na, ⁶⁰Co, and ⁵⁶Mn.

The South Atlantic Anomaly (SAA) was modeled using the AP9 trapped‑proton spectrum (IRENE model). The proton flux above 4 MeV was integrated over each 15‑s interval to generate a light curve representing SAA passages (average duration ≈17.5 min). The authors deliberately omitted trapped electrons because their energies are too low to cause significant spallation. To validate the AP9/AE9 predictions, they compared them with measurements from the High‑Energy Particle Detector (HEPD‑01) aboard the China Seismo‑Electromagnetic Satellite (CSES‑01) at a similar altitude (≈500 km) and low latitude (|b| ≤ 1°). The comparison revealed that AP9/AE9 overestimates the trapped‑proton flux by a factor of ~9, confirming earlier indications that these models are too high for low‑inclination, low‑altitude orbits.

Simulation results show a clear energy dependence of the dominant background component. Below ~660 keV the extragalactic photon background dominates, contributing >70 % of reconstructed Compton events. Between 660 keV and ~2 MeV, prompt interactions of primary CR protons/α‑particles and delayed activation from short‑lived isotopes become the main sources. Above ~2 MeV, delayed activation from long‑lived isotopes produced by CR interactions and albedo photons takes over. The authors quantify systematic uncertainties: ≈10 % from input spectra, ≈5 % from Rcut variations, and ≈20 % from activation modeling, with the SAA model discrepancy contributing an additional ≈5 % to the total background budget.

Overall, the paper delivers a comprehensive pre‑flight background model for COSI, identifies the relative importance of each component across the instrument’s energy band, and highlights the need to correct the AP9/AE9 trapped‑proton predictions for low‑inclination missions. These results will feed directly into the COSI data‑analysis pipeline, enabling more accurate event selection, imaging reconstruction, and sensitivity estimates throughout the mission.