Forecasting Space Weather

Forecasting space weather events presents the ultimate challenge to a space physics model. Not only should physical constraints be satisfied, but also practical issues such as timeliness, accuracy, and reliability must be adequately addressed. To address these needs, modern space weather forecasters and users rely on a great variety of space weather models, ranging from simple regressions and ending with fairly complex information-based (empirical), physical, and hybrid models. As a result, model-based predictions of space environments have steadily improved. Especially since the early 1990s when real-time data started being made available on-line, time-dependent inputs, and later data assimilation and related techniques, such as Kalman filtering, have significantly improved prediction accuracy. Model validation and verification are important steps in assessing a space weather model's full spectrum of capabilities. I discuss the above concepts and illustrate them where necessary using case studies. I conclude by summarizing relevant recent developments and future outlook.-3