INTEGRAL Picture Of the Month
March 2022

INTEGRAL POM
(Click to download full resolution)

INTEGRAL supersedes COMPTEL in measuring the Galactic diffuse MeV emission

For more than 20 years, the Compton telescope, COMPTEL, onboard NASA's Compton Gamma-Ray Observatory, CGRO, set the standard for measurements of the extended emission of the Milky Way in the MeV range. This photon energy range harbours many emission mechanisms, for example, the annihilation of positrons in the interstellar medium, nuclear decays of radioactive isotopes, or the Inverse Compton scattering of electrons off the interstellar radiation field. These components are indicated in the top panel, showing the diffuse Galactic emission spectrum with its characteristic spectral features [1].

Cosmic-ray GeV electrons scatter off photons from star light in the optical and infrared (IR), as well as the cosmic microwave background (CMB). This scattering produces a smooth spectrum which peaks around 0.1 MeV for the CMB, around 0.5 MeV for IR, and potentially above 10 MeV for optical light. Together this makes the smooth spectrum shown in the zoomed-in panel [2]. Measuring the range between 0.5 MeV up to the limit of INTEGRAL/SPI at 8 MeV therefore constrains the transport properties of electrons as they scatter in IR and optical light.

With more than 16 years of data, SPI was able to set a new record in determining the diffuse emission spectrum in the 0.5-8.0 MeV band. The four coloured data points shown have a signal-to-noise ratio of 6 or more, superseding the data quality and precision of COMPTEL (green). The projections of the measured photons above the strong instrumental background [3] is shown for the four data points surrounding the bottom panel: there is clearly flux originating from the Galactic plane.

In this study, it was possible to test extreme assumptions on the propagation of GeV electrons: it appears that one common diffusion coefficient of 0.5 for the entire cosmic-ray electron spectrum best matches the INTEGRAL/SPI data. Another important finding in the work is that the absolute normalisation of most propagation models is a factor of 2-3 below the measurements, opening up the possibility of even more emission components such as unresolved point sources.

References:

back to the POM archive