Will it rain tomorrow? Improving probabilistic forecasts

Predicting the weather is notoriously difficult because of its chaotic nature. In chaotic systems, small errors in the initial conditions can cause large errors in predicting the future. In this poster we describe the challenges of such systems and a method that has been developed to deal with the uncertainty it brings. Ensemble forecasting uses a variety of initial conditions that take into account any uncertainty in what is happening now to give us a variety of predictions of the future, each accounting for different circumstances. We compare a variety of different methods of transforming these ensembles into probability distributions which can be of great use to decision makers.