The forecast skill horizon

Abstract This chapter reviews the progress in numerical weather prediction, introduces the forecast skill diagram and the concept of the forecast skill horizon, and discusses predictability issues. More specifically, it addresses the following questions: (1) Are weather forecasts more accurate and reliable today than in the past? (2) How did we manage to improve the accuracy and reliability of weather forecasts? (3) Can we visualize in a single diagram our prediction capabilities? (4) Why does the forecast skill depend on the phenomena we are trying to predict? (5) Are extreme events more difficult to predict than the “normal” weather? (6) What is the minimum spatial scale that a model can realistically simulate? (7) What is the minimum spatial scale properly resolved in data assimilation? And (8) How can we further extend predictability?