Climate change remains a defining issue of our time, and there is a pressing requirement for highly accurate estimates of expected alteration to global and regional weather patterns. Unfortunately, even the most recent set of Earth System Model (ESM) projections are widely divergent. The DPhil student will use the latest powerful mathematical, statistical and AI methods, to merge climate measurements with ESM projections, and reduce model spread. The focus will be on tropical climate features, and with a particular emphasis on interannual variability. If time available, the findings will be converted to impacts, and for instance, by forcing land surface models with predicted climate applicable for different atmospheric greenhouse gas concentrations.