A new look at the cosmic ray positron fraction

Context. The positron fraction in cosmic rays has recently been measured with improved accuracy up to 500 GeV, and it was found to be a steadily increasing function of energy above ~10 GeV. This behaviour contrasts with standard astrophysical mechanisms, in which positrons are secondary particles, produced in the interactions of primary cosmic rays during their propagation in the interstellar medium. The observed anomaly in the positron fraction triggered a lot of excitement, as it could be interpreted as an indirect signature of the presence of dark matter species in the Galaxy, the so-called weakly interacting massive particles (WIMPs). Alternatively, it could be produced by nearby sources, such as pulsars. Aims: These hypotheses are probed in light of the latest AMS-02 positron fraction measurements. As regards dark matter candidates, regions in the annihilation cross section to mass plane, which best fit the most recent data, are delineated and compared to previous measurements. The explanation of the anomaly in terms of a single nearby pulsar is also explored. Methods: The cosmic ray positron transport in the Galaxy is described using a semi-analytic two-zone model. Propagation is described with Green functions as well as with Bessel expansions. For consistency, the secondary and primary components of the positron flux are calculated together with the same propagation model. The above mentioned explanations of the positron anomaly are tested using χ2 fits. The numerical package MicrOMEGAs is used to model the positron flux generated by dark matter species. The description of the positron fraction from conventional astrophysical sources is based on the pulsar observations included in the Australia Telescope National Facility (ATNF) catalogue. Results: The masses of the favoured dark matter candidates are always larger than 500 GeV, even though the results are very sensitive to the lepton flux. The Fermi measurements point systematically to much heavier candidates than the recently released AMS-02 observations. Since the latter are more precise, they are much more constraining. A scan through the various individual annihilation channels disfavours leptons as the final state. On the contrary, the agreement is excellent for quark, gauge boson, or Higgs boson pairs, with best-fit masses in the 10 to 40 TeV range. The combination of annihilation channels that best matches the positron fraction is then determined at fixed WIMP mass. A mixture of electron and tau lepton pairs is only acceptable around 500 GeV. Adding b-quark pairs significantly improves the fit up to a mass of 40 TeV. Alternatively, a combination of the four-lepton channels provides a good fit between 0.5 and 1 TeV, with no muons in the final state. Concerning the pulsar hypothesis, the region of the distance-to-age plane that best fits the positron fraction for a single source is determined. Conclusions: The only dark matter species that fulfils the stringent gamma ray and cosmic microwave background bounds is a particle annihilating into four leptons through a light scalar or vector mediator, with a mixture of tau (75%) and electron (25%) channels, and a mass between 0.5 and 1 TeV. The positron anomaly can also be explained by a single pulsar, and a list of five pulsars from the ATNF catalogue is given. We investigate how this list could evolve when more statistics are accumulated. Those results are obtained with the cosmic ray transport parameters that best fit the B/C ratio. Uncertainties in the propagation parameters turn out to be very significant. In the WIMP annihilation cross section to mass plane for instance, they overshadow the error contours derived from the positron data.