In this session, Hamish Steptoe from the MET Office explores how conflicting datasets can lead to policy inaction – and how data science offers a path through. Using summer rainfall extremes in Nepal as a case study, we reveal how common open-access datasets can produce wildly different results. We’ll share how data blending and statistical modelling techniques can reduce uncertainty, improve confidence, and support more informed climate-related decision-making at national and regional levels.