First in a series on probabilistic forecasting for physical signals. Next: what happens when you roll the forecast forward more than one step.