<p>In this study, the statistical characteristics and mechanisms of the Southern Indian Ocean Dipole (SIOD) from 1979 to 2023 are examined using sea surface temperature (SST) data from the Hadley Centre, reanalysis data from the NCEP/NCAR, and numerical experiments conducted with the Community Earth System Model (CESM). In boreal winter, the first mode derived from empirical orthogonal function analysis of sea surface temperature anomalies (SSTAs) in the Southern Indian Ocean reveals a dipole mode oriented in the southwestern–northeastern direction, referred to as the SIOD. The SIOD exhibits short-period oscillations of 2–3&#xa0;years and long-period oscillations of 6–7&#xa0;years, typically developing in boreal autumn, peaking in winter, and decaying in the following spring. The mechanisms of the SIOD are jointly influenced by both dynamic and thermodynamic factors, including wind speed, latent heat flux, subtropical high, and mixed layer depth. In addition to the well-known influence of the Mascarene High, the role of the Australian High (AH) should not be ignored. Specifically, a weakened AH promotes the development of a positive SIOD [positive SSTAs in the southwestern pole (SWP), negative SSTAs in the northeastern pole (NEP)], while a strengthened AH favors a negative SIOD (negative SSTAs in the SWP, positive SSTAs in the NEP). Numerical simulations using CESM effectively reproduce the anomalous changes in sensible heat flux and latent heat flux during the autumn and winter of both positive and negative SIOD event years. It is particularly noteworthy that the SSTAs in the SWP lag behind those in the NEP in the autumn of positive SIOD event years. Furthermore, the center of positive SSTAs shows a clear eastward shift in the following spring. In contrast, the SSTAs of the SWP and NEP develop synchronously during negative SIOD events, so their occurrence time is earlier than that of the positive SIOD events. This asymmetry can be attributed to differences in key driving factors.</p>