Models for Galactic Cosmic-ray Propagation

A new numerical model of particle propagation in the Galaxy has been developed, which allows the study of cosmic-ray and gamma-ray production and propagation in 2D or 3D, including a full reaction network. This is a further development of the code which has been used for studies of cosmic ray reacceleration, Galactic halo size, antiprotons and positrons in cosmic rays, the interpretation of diffuse continuum gamma rays, and dark matter. In this paper we illustrate recent results focussing on B/C, sub-Fe/Fe, ACE radioactive isotope data, source abundances and antiprotons. From the radioactive nuclei we derive a range of 3-7 kpc for the height of the cosmic-ray halo. 0 200 1 COSPAR. Published by Elsevier Science Ltd. All rights reserved. INTRODUCTION In recent years, new and accurate data have become available in cosmic-ray (CR) astrophysics, and much progress in understanding the origin and propagation of CR can be made, but this requires a systematic and self-consistent approach. With this motivation we have developed a numerical propagation code which aims to reproduce observational data of many kinds related to CR, such as direct measurements of nuclei, antiprotons, positrons and electrons, gamma rays and synchrotron radiation. In the first version, the model was cylindrically symmetric and was written in fortran-90. This version has proven that the original goal, a “realistic standard model” of CR production and propagation, is possible to achieve. It has been used for studies of CR reacceleration, halo size, production and propagation of positrons and antiprotons, dark matter annihilation, and the interpretation of diffuse continuum gamma rays. Many aspects, however, cannot be addressed in such a model, e.g. the stochastic nature of the CR sources in space and time which is important for high-energy electrons with short cooling times, and local inhomogeneities in the gas density which can affect radioactive secondary ratios in CR. The experience gained from the original version allowed us to design a new version of the model, entirely rewritten in C++, which is much more flexible, which incorporates essential improvements over the older model, and in which a S-dimensional spatial grid can be employed. In addition it is now possible to solve the full nuclear reaction network on the spatially resolved grid. We keep however an option which allows us to switch to the older cylindrically symmetrical (“2D”) model since this is still a sufficient approximation for nuclei in the absence of spatial fluctuations and is much faster to compute than the full 3D case. The code can thus serve as a complete substitute for the conventional “leaky-box” or “weighted-slab” propagation models usually employed, with many associated advantages such as the correct treatment of radioactive nuclei, realistic gas and source distributions etc. The present work illustrates the use of the new 2D code to interpret B/C, sub-Fe/Fe, new radioactive isotopic ratio data from ACE, source abundances and antiprotons. *NRC Smior Research Associate ‘cm leave from Institute of Nuclear Physics, M.V.Lornonosov Moscow State University, Moscow 119899, Russia 718 A. W. Strong and I. V Moskalenko Elemental Abundances 0 200 MeWnucleon Fig. 1. Relative CR propagated elemental abundances as calculated in our model at 200 MeV/nucleon in the heliosphere circles. Modulation potential 9 = 450 MV. Data: Duvernois and Thayer (1996) vertical bars.