Everyone has an interest in the weather, whether it's to check the prospects for a day out or to know when best to harvest a crop. Until relatively recently, weather forecasts could be wildly inaccurate. This was due to the mathematically chaotic nature of weather systems, first discovered in the 1960s, understanding of which would transform forecasting from the 1990s and mean that meteorologists became amongst the foremost users of supercomputers. From the crystalline perfection of the snowflake to transfer of energy from the Sun, science lies at the heart of the weather and our understanding of it. In recent years, weather science has moved to the leading edge with advanced modelling, versatile use of satellite data and a better understanding of mathematical chaos. This is a true example of hot science at work.