(1) Data Management at IMK-TRO (2) Observational study of land-lake breeze over Lake Victoria basin in Uganda using station data (3) A neural network for calculating optical properties of internally mixed aerosol in ICON-ART (3) tbd (4) Illuminating convective cold pools with a dense station network

In this wide-ranging talk, I will introduce machine learning and the many ways in which we use it at the Institute for Climate and Atmospheric Sciences (ICAS). I'll touch upon some of the reasons that it is generating excitement in weather forecasting and introduce some of the deeper debates about the extent to which such methods are interpretable. I'll also delve into the practical side of how we (try) to keep up with the never-ending generation of new algorithms and techniques.