The Internal and ENSO-Forced Modes of the Indian Ocean Sea Surface Temperature

The internal and ENSO-forced modes of the Indian Ocean sea surface temperature (SST) are investigated using a high-resolution regional coupled model. Five different model simulations were performed by controlling atmospheric and oceanic boundary conditions (BCs), which are lateral walls of the model domain. In the internal run performed by prescribing the climatological mean oceanic and atmospheric BCs, the first and second empirical orthogonal functions (EOF1 and EOF2) of internal mode are similar to the observed Indian Ocean basin (IOB) and dipole (IOD) modes with relatively weak amplitudes, respectively. In the control run with observed BCs, those EOFs are much amplified with their power spectrums significantly changed, and their spatial patterns are modified, particularly for the EOF2. Three ENSO runs with combinations of ENSO-related and climatological mean BCs show that the modification of spatial pattern of EOFs is mainly due to ENSO forcing. Furthermore, ENSO forcing determines the major 4-yr period of IOB mainly through the atmosphere, whereas the major 3-yr period of IOD is determined by both ENSO atmosphere and ocean forcings. It is also found that IOB and IOD exhibit a significant seasonally dependent relationship in both internal and ENSOforced simulations. Most importantly, by applying the empirical singular vector method to both observed and modeled data, it is found that the IOD-IOB relationship is associated with an unstable mode of Indian Ocean SST anomalies, evolving from boreal fall to the next spring. This unstable mode is intrinsic within the Indian Ocean but is substantially amplified by the ENSO.

Automated summary

This study investigates the internal and ENSO-forced modes of Indian Ocean sea surface temperature (SST) using a high-resolution regional coupled model. The study finds that ENSO forcing is the main factor that modifies the spatial patterns of EOFs, and ENSO also determines the major periods of IOB and IOD. Additionally, the study discovers a seasonally dependent relationship between IOB and IOD in both internal and ENSO-forced simulations.