Total: 1
This study examines the impact of solar modulation on the antiproton excess observed by AMS-02, which may indicate dark matter (DM) annihilation. We analyze three solar modulation models: the force-field approximation (FFA), a time-, charge-, and rigidity-dependent FFA, and a three-dimensional numerical simulation based on the Parker transport equation. Based on AMS-02 latest antiproton data (2025), our results show that the significance of the DM signal is sensitive to the chosen modulation model, with a 2$\sigma$ signal for the FFA (4$\sigma$ if including data from H, He, C, O, B/C, and B/O) and a reduced significance for more complex models. We also address systematic uncertainties using two methods: the add-in-quadrature method, which assumes uncorrelated uncertainties between energy bins, and the nuisance parameter method, which treats systematic uncertainties as nuisance parameters during the fitting process. Fitted to AMS-02 antiproton data, DM annihilation to the $b\bar{b}$ scenario with three different solar modulation models shows that the add-in-quadrature method causes overfitting, whereas the nuisance parameters approach leads to underfitting. Statistically, the signal region of the FFA model using the add-in-quadrature method is the most reliable. This work highlights the need for refined solar modulation models and a better treatment of uncertainties for a conclusive interpretation of the AMS-02 data.