Attribution of Interannual Variability of the Upper- and Lower-Layer Indonesian Throughflow Based on Adjoint Sensitivity

Abstract

The Indonesian Throughflow (ITF) and its variability significantly impact regional oceanography and global climate system. To attribute ITF interannual variability to local and remote forcing, a near-global ocean model and its adjoint are utilized to derive adjoint sensitivities of upper-layer (<300 m) and lower-layer (>300 m) ITF to surface momentum and buoyancy fluxes with spatiotemporal details. The historical ITF simulation aligns closely with reanalysis data and observational data in both layers. The adjoint sensitivity reveals that ITF interannual variability is primarily driven by wind stress occurring within the preceding 6 months, while buoyancy flux contributions are negligible. Combined with perturbation experiments, we demonstrate the physical interpretations of the derived adjoint sensitivities. The near-opposite adjoint sensitivities between upper- and lower-layer ITF suggest that wind stress impacts the ITF mainly through wind-driven currents, Kelvin wave and Rossby wave propagation along the equatorial Indo-Pacific region. Specifically, the lower-layer ITF is primarily influenced by wind stress over the equatorial Indian Ocean, which generates opposite variabilities in the upper layer. Therefore, the equatorial Pacific emerges as the dominant driver of full-depth ITF interannual variability, with a strong El Niño–Southern Oscillation (ENSO) signature. This study demonstrates that the robust relationship between ITF interannual variability and wind stress changes is independent of both ENSO and seasonal variations. As a novel application, this approach enables historical reconstruction of the interannual variability for both the full-depth ITF and its vertical structure using reliable wind data.