A Hybrid Gravest Empirical Mode Method for Reconstructing Temperature Profiles in the Seychelles–Chagos Thermocline Ridge

The acoustic travel time (τ) measured by an inverted echo sounder (IES) can be converted into vertical temperature profiles using the gravest empirical mode (GEM) technique based on the relationship between τ and temperature. In the Seychelles–Chagos Thermocline Ridge (SCTR) in the southwestern tropical Indian Ocean, where persistent subsurface upwelling occurs, continuous vertical temperature profiles are crucial for monitoring the upwelling variability. However, the conventional GEM method has large uncertainties at the SCTR due to temporal variability in upwelling strength. This study introduces a new approach, termed hybrid GEMs, to improve IES data analysis by reflecting upwelling strength based on the depth of the 20°C isotherm (D20). The hybrid GEMs consist of one moderate GEM and two combined GEMs derived from three groups of historical hydrographic profiles in the SCTR, categorized by D20 ranges. When applied to the in situ τ measured by a pressure-recording IES at Station K (61°E, 8°S) in the SCTR from May 2019 to December 2021, the absolute dynamic topography from satellite altimetry is used as an index to select the appropriate hybrid GEM based on the consistency between the absolute dynamic topography and D20 variability. The vertical temperature profiles based on hybrid GEMs show significant improvements in both the mean and maximum root mean square errors of the upper 300 m temperature, which are reduced by approximately 29% and 20%, respectively. The hybrid GEM–derived temperature profiles reliably capture temperature variability in the upper 300 m, demonstrating the strong potential of acoustic travel time as an essential observational variable in data-sparse tropical upwelling regions of the Indian Ocean.