This manuscript examined sea level simulation in the tropical Pacific by the GFDL-OM4 ocean model driven by two different atmospheric products: CORE and JRA55do. Long-term mean, linear trends, seasonal and interannual variability are discussed in detail, in particular focusing on the differences between model results and observations, and between two model experiments. It's found that the JRA55-do tends to give improved simulation, closer to observations than CORE. C1 GMDD Interactive comment Printer-friendly version Discussion paper