Modelling and forecasting the Indian Ocean Dipole

Gler, M. (2026) Modelling and forecasting the Indian Ocean Dipole. PhD thesis, University of Reading. doi: 10.48683/1926.00131151

Abstract/Summary

The Indian Ocean strongly influences climate variability in surrounding countries, particularly through its dominant coupled mode of interannual variability, the Indian Ocean Dipole (IOD). However, current climate and seasonal forecast models struggle to simulate the Indian Ocean mean state and its interannual variability. This thesis evaluates models from the sixth phase of the Coupled Model Intercomparison Project (CMIP6) and the Met Office seasonal forecasting system, GloSea6, focusing on their representation of the Indian Ocean and the IOD. Analysis of atmosphere-only and coupled CMIP6 models, including those from the Met Office, reveals a common dipole pattern of mean-state atmospheric biases across the Indian Ocean during boreal summer and autumn. Further analysis shows that ocean-atmosphere coupling exacerbates these biases, with coupled models simulating a stronger-than-observed IOD amplitude and regional precipitation teleconnections. The development of the persistent coupled bias is further examined in GloSea6, which shares the same configuration as the Met Office’s CMIP6 models and exhibits a similar bias pattern. Initialised hindcasts in GloSea6 show that the bias emerges over the eastern equatorial Indian Ocean in summer, with an easterly wind bias likely linked to the summer monsoon circulation, and persists into autumn over the western equatorial Indian Ocean via the Bjerknes feedback. GloSea6 shows high IOD prediction skill, at lead times of up to two months, but overestimates its amplitude. To further diagnose the sources of the coupled bias, atmospheric nudging experiments are performed using a Met Office CMIP6 model, in which winds and temperatures are relaxed towards reanalysis over global and specific Indian Ocean domains. Results show a reduction in the magnitude of atmospheric and oceanic biases, and an improved simulation of IOD variability. This study demonstrates that reducing Indian Ocean regional atmospheric biases in boreal summer and autumn is essential for improved representation of its climate variability and teleconnections.

Item Type	Thesis (PhD)
URI	https://centaur.reading.ac.uk/id/eprint/131151
Identification Number/DOI	10.48683/1926.00131151
Divisions	Science > School of Mathematical, Physical and Computational Sciences > Department of Meteorology
Download/View statistics	View download statistics for this item