Indonesian Throughflow in an eddy-resolving ocean model

The performance of the eddy-resolving LICOM2.0 in simulating the Indonesian Throughflow has been evaluated against the INSTANT data in the present study. The mean vertical structures of the along strait velocities are simulated well in LICOM2.0, but the large velocities at the bottom of the Lifamatola Passage and the Timor Passage cannot be reproduced by LICOM2.0. The causes are considered to be both the errors in the topography and the tidal mixing at the bottom. Despite several biases in the mean velocities, the mean inflow and outflow volume transports in LICOM2.0 are almost identical to the INSTANT data. Compared with the lower resolution LICOM, the most significant improvement is the better simulation of the partitions of the inflow and outflow transports in individual straits. The outflow for low-resolution LICOM is mainly through the Ombai and Lombok Strait, whereas that for LICOM2.0 is mainly through the Timor Passage. The variability of the vertical structure of velocities and the volume transport are also investigated. LICOM2.0 overestimates the magnitude of the upper-layer currents and the amplitude of the variation. We also found that the largest correlation coefficient occurs in the shallowest strait, the Lombok, whereas the lowest occurs in the Timor Passage, especially in the upper layer. The latter may be caused by the unrealistic transport through the Torres Strait in LICOM2.0.