The tropical Indian Ocean has experienced pronounced warming trends in recent decades, with dynamical processes recognized as key drivers. However, the role of thermodynamic processes remains uncertain due to discrepancies in surface wind-induced heat flux across existing datasets. Here, we utilize a machine learning algorithm to integrate in-situ observations and satellite data, yielding a reliable surface wind dataset and corresponding air-sea heat flux spanning from 1950–2019 with a horizontal resolution of 1°×1°. Our analysis reveals a weakening of surface wind over the tropical Indian Ocean since 1950, supported by variations in sea surface height and thermocline depth. Consequently, thermodynamic processes associated with surface wind-induced heat flux promote warming in the eastern tropical Indian Ocean, accounting for 45% of the contributions of dynamical processes. These findings challenge reanalysis results but are aligned with state-of-the-art models, underscoring that the significance of thermodynamic processes is substantially underestimated by current reanalysis datasets.

Indian Ocean warming, surface wind, machine learning