Electron capture (EC) isotopes are known to provide constraints on the low energy behavior of cosmic rays (CRs), such as re-acceleration. Here we study the EC isotopes within the framework of the dynamic spiral- arms CR propagation model in which most of the CR sources reside in the galactic spiral arms. The model was previously used to explain the B/C and sub-Fe/Fe ratios (Benyamin et al. 2014, 2016). We show that the known inconsistency between the 49Ti/49V and 51V/51Cr ratios remains also in the spiral-arms model. On the other hand, unlike the general wisdom in which the isotope ratios depend primarily on reacceleration, we find here that the ratio also depends on the halo size (Zh) and in spiral-arms models also on the time since the last spiral arm passage (τarm). Namely, EC isotopes can in principle provide interesting constraints on the diffusion geometry. However, with the present uncertainties in the lab measurements of both the electron attachment rate and the fragmentation cross-sections, no meaningful constraint can be placed.

Most of the diffuse Galactic GeV γ-ray emission is produced via collisions of cosmic ray (CR) protons with ISM protons. As such the observed spectra of the γ-rays and the CRs should be strongly linked. Recent observations of Fermi-LAT exhibit a hardening of the γ-ray spectrum at around a hundred GeV, between the Sagittarius and Carina tangents, and a further harden- ing at a few degrees above and below the Galactic plane. However, standard CR propagation models that assume a time independent source distribution and a location independent diffu- sion cannot give rise to a spatially dependent CR (and hence γ-ray) spectral slopes. Here we consider a dynamic spiral arm model in which the distribution of CR sources is concentrated in the (dynamic) spiral arms, and we study the effects of this model on the π0-decay produced γ-ray spectra. Within this model, near the Galactic arms the observed γ-ray spectral slope is not trivially related to the CR injection spectrum and energy dependence of the diffusion co- efficient. We find unique signatures that agree with the Fermi-LAT observations. This model also provides a physical explanation for the difference between the local CR spectral slope and the CR slope inferred from the average γ-ray spectrum.

The boron to carbon (B/C) and sub-Fe/Fe ratios provide an important clue on cosmic ray (CR) propagation within the Galaxy. These ratios estimate the grammage that the CRs traverse as they propagate from their sources to Earth. Attempts to explain these ratios within the standard CR propagation models require ad hoc modifications and even with those these models necessitate inconsistent grammages to explain both ratios. As an alternative, physically motivated model, we have proposed that CRs originate preferably within the galactic spiral arms. CR propagation from dynamic spiral arms has important imprints on various secondary to primary ratios, such as the B/C ratio and the positron fraction. We use our spiral-arm diffusion model with the spallation network extended up to nickel to calculate the sub-Fe/Fe ratio. We show that without any additional parameters the spiral-arm model consistently explains both ratios with the same grammage, providing further evidence in favor of this model.

We develop a fully three-dimensional numerical code describing the diffusion of cosmic rays (CRs) in the Milky Way. It includes the nuclear spallation chain up to oxygen, and allows the study of various CR properties, such as the CR age, grammage traversed, and the ratio between secondary and primary particles. This code enables us to explore a model in which a large fraction of the CR acceleration takes place in the vicinity of galactic spiral arms that are dynamic. We show that the effect of having dynamic spiral arms is to limit the age of CRs at low energies. This is because at low energies the time since the last spiral arm passage governs the CR age, and not diffusion. Using the model, the observed spectral dependence of the secondary to primary ratio is recovered without requiring any further assumptions such as a galactic wind, re-acceleration or various assumptions on the diffusivity. In particular, we obtain a secondary to primary ratio which increases with energy below about 1 GeV.

Cosmic Ray diffusion models lacking nearby sources require a smaller halo and smaller diffusion coefficient in order to fit the average grammage inferred from ratios between secondary to primary cosmic rays as well as the 10Be/9Be which “dates” the cosmic rays. We show here that models with such modified parameters lead to a notably smaller anisotropy, releasing the tension between standard predictions and observations. For the lower diffusion coefficient, the “knee” at a few PeV is then unavoidably explained as the energy for which the Larmor radius is equal to the mean free path, because of which the ISM diffusion coefficient will increase fast above this energy (for protons). This would imply that the knee is a propagation effect and the main cosmic ray acceleration site (presumably supernova remnants) should in principle be able to accelerate up to higher energies. The smaller scales also imply that fewer sources contribute to the cosmic ray density at any given energy, giving rise to large fluctuations which can explain the anisotropy behavior above 10 TeV. Although the energy dependent composition is similar to the behavior expected in standard models, here one predicts a larger anisotropy for lighter elements at a given energy above the knee.

We have previously focused on studying the electron-capture isotopes within the dynamic spiral-arms model and empirically derived the energy dependence of the electron attachment rate using the observation of 49Ti/49V and 51V/51Cr ratios (Benyamin et al. 2017). We have also shown how this relation recovers the energy dependence seen in the lab measurements (Letaw et al. 1985). In this work we use this relation to construct the 44Ca/44Ti ratio and place a lower limit on the amount of 44Ti that is required to be nucleosynthesized at the source. The results also imply that the acceleration process of the radioisotopes cannot be much longer than a century time scale (or else the required nucleosynthesized amount has to be correspondingly larger). We also provide a similar lower limit on the source 60Fe by comparing to the recently observed 60Fe/56Fe (Binns et al.
2016).