Abstract The Equatorial Undercurrent (EUC) in the Indian Ocean (IO) is a key component of upper‐ocean circulation, exhibiting pronounced interannual asymmetry during boreal autumn, with eastward (positive EUC; pEUC) anomalies being stronger than westward (negative EUC; nEUC) anomalies. This study assesses the performance of Coupled Model Intercomparison Project Phase 6 (CMIP6) models in capturing the skewness of the autumn EUC. While a subset of CMIP6 models (positive‐skewed group) reasonably reproduces the observed positive skewness, the majority underestimate it, and several (negative‐skewed group) simulate a reversed, negative skewness. The results reveal that these discrepancies are closely linked to the mean‐state biases resembling a positive Indian Ocean Dipole (pIOD). Models with weaker pIOD‐like biases exhibit a deeper climatological thermocline and warmer sea surface temperatures (SSTs) in the eastern IO, conditions that favor more robust thermocline shoaling‐induced cooling relative to deepening‐induced warming. This asymmetric thermocline–temperature feedback enhances surface‐subsurface coupling among the thermocline depth, SSTs, and zonal wind stress, thereby amplifying asymmetries in subsurface pressure gradient forces (PGFs). Both the sea surface height‐induced PGFs, mediated by equatorial wave dynamics, and the density‐induced PGFs, driven by subsurface temperature gradients, contribute to the observed EUC skewness. However, models with stronger pIOD‐like biases simulate a shallower thermocline and cooler eastern IO SSTs, which suppress wind‐driven thermocline responses and weaken surface‐subsurface coupling, ultimately reducing PGF asymmetry and EUC skewness. These findings highlight the importance of correcting mean‐state biases in coupled climate models to improve the simulation of equatorial ocean dynamics and their variability.
Li et al. (2026) studied this question.