Recent precise measurements of cosmic ray spectra revealed an anomalous hardening at ∼200 GV, observed by the ATIC, CREAM, PAMELA, and AMS02 experiments. Particularly, the latest observation of the ratio by AMS02 demonstrated a flat distribution, which further validated the spectral anomalies of secondary particles. All those new phenomena indicated that the conventional propagation model of cosmic rays meets a challenge. In this work, the spatial-dependent diffusion coefficient is employed by tracing the source distribution under the physical picture of the two-halo model in the DRAGON package. Under such a scenario, the model calculation will result in two-component spectra for primary nuclei. Due to the smaller rigidity dependence of in the galactic disk, the ratios secondary-to-primary will inevitably be flatter and the expected anisotropy of cosmic rays will be much more attenuated than in the conventional model. As a result, we can reproduce the spectral hardening of protons, the flat ratios of and B/C, and consistent anisotropy from ∼100 GeV to ∼100 TeV by only adopting one set of spatial-dependent diffusion coefficients in a galactic disk.

Recently the AMS-02 collaboration published detections of light cosmic-ray nuclei, including lithium, beryllium, boron, carbon, and oxygen. Combined with the released energy spectra of proton and helium, both primary and secondary spectra have a prominent hardening above ∼200 GV. In particular, the spectral variation of secondary cosmic rays is greater than the primary ones. One of the plausible interpretations for the above anomalies is the spatial-dependent diffusion model. It has successfully described various observational phenomena, e.g., hardening of primary nuclei, diffuse gamma-ray distribution and cosmic-ray anisotropy, etc. In this work, we apply the spatial-dependent propagation model to the latest observations, including both primary and secondary nuclei. Apart from the primary components, the spectra of secondary nuclei can be properly reproduced as well, especially the upturn above 200 GV. We also calculate the ratios of both secondary-to-primary and secondary-to-secondary. We find that except for the Be/B ratio, the computations of the spatial-dependent propagation model are in good agreement with the current data.