Understanding the South-West Indian Ocean Temperature Bias: Forcing Ocean-Only Models with Coupled Model Atmospheric Output to Identify Drivers of Ocean Biases

Abstract

The Seychelles-Chagos Thermocline Ridge (SCTR) is a region of open-ocean upwelling in the south-west tropical Indian Ocean. It plays a prominent role in regional ocean dynamics, local and remote climate variability, and can influence the onset of the Indian summer monsoon and the formation of tropical cyclones in the region. This highlights the importance for global coupled models to be able to accurately simulate the dynamics of the SCTR. The upwelling in the SCTR is primarily wind-driven, namely by the local negative wind stress curl between the south-easterly trade winds and the equatorial westerlies. Therefore, any model biases present in the local winds will have a subsequent effect on the strength of the upwelling. The Met Office global coupled model exhibits a significant subsurface warm bias in the SCTR region.  Its location makes it part of the western pole of the Indian Ocean dipole; therefore, its existence affects the representation of Indian Ocean climate modes and their associated teleconnections. The bias shows a strong seasonal cycle, peaking in boreal winter and subsiding from spring. This warm bias is not seen in the ocean-only model when forced by reanalysis data. We hypothesise that the temperature bias is due to a lack of ocean upwelling caused by biases in the surface winds in the coupled model. To test this hypothesis, an ocean-only model was forced by the surface winds from the coupled model with all other atmospheric forcings taken from reanalysis. We also test the impact of forcing the ocean model with winds from an atmosphere only simulation. We will examine whether the coupled and atmosphere-only wind biases are effective in recreating the temperature bias in the ocean-only model and investigate the potential role of coupled feedbacks.