Systematic biases over the equatorial Indian Ocean and their influence on seasonal forecasts of the IOD

Accurate representation and predictability of the Indian Ocean Dipole (IOD) in seasonal forecasts are crucial given its pronounced socioeconomic impacts on countries surrounding the Indian Ocean.  Using hindcasts from the coupled Met Office Global Seasonal Forecasting System (GloSea6), coupled mean state biases in the western and eastern equatorial Indian Ocean (WEIO and EEIO) and their impacts on IOD prediction are examined. Results show that GloSea6 exhibits a pronounced cold bias in the EEIO that rapidly develops after the onset of the monsoon in boreal summer (JJA, July-August) and persists through the autumn season (SON, September-November).  This cold bias, along with a dry bias, is linked to erroneous easterlies and a shallow thermocline, likely associated with the summer monsoon circulation.  The seasonal evolution and relative timing of the precipitation biases between the western and eastern IOD poles, such that the biases develop through JJA in the EEIO but follow in the WEIO in SON, suggests that the EEIO plays the leading role in the development of coupled feedbacks that result in an overall large dipole pattern of atmospheric and subsurface oceanic biases in SON.Analysis of skill metrics for the IOD shows that GloSea6 achieves a high anomaly correlation coefficient at short lead times, though it tends to overestimate IOD ampltiude, indicating higher IOD variability compared to observations.  This overestimation is larger in the eastern IOD pole than in the western pole and is likely linked to the poor representation of the evolution of the SST anomalies in the EEIO during positive and negative IOD events in SON.  This aligns with the skill metrics of the individual poles, which show a lower anomaly correlation coefficient and higher prediction errors observed in the eastern pole compared to the west.Results in this study highlight the crucial role of regional biases, particularly in the EEIO, in shaping IOD variability and suggest that addressing these regional biases in GloSea6 could improve IOD prediction skill, enhancing forecasts of climate impacts for countries surrounding the Indian Ocean.