Predicting and Modeling Space Weather

Abstract:

Predicting space weather on Earth is a maturing, but still inexact
science. Just as meteorologists follow the path of hurricanes and try
to predict the time, location and intensity of landfall, the longer
ahead the prediction, the less accurate the prediction becomes.
Similarly, predicing the time of a CME arrival at Earth is still not
very well modeled by solar wind stream modelers, and predicting the
north/south component of the IMF from solar measurements (which
affects the intensity of the resulting space weather) is just now
being successfully attempted. Predicting space weather effectsfrom near-earth monitors such as ACE and now DISCOVR, is much more accurate but has a relatively short effective prediction time (1-3 hours).

Our solar wind forecast system, which uses ACE data, has not missed a major storm in 12 years (http://mms.rice..edu/realtime/forecast.html). We have extended our forecast to six days by combining with ENLIL, but with considerably less accuracy. For the best space weather predictions, one would like to have a monitor even farther upstream than L1, far enough to give a longer prediction time frame (4-6 hours), but not so far that the solar wind correlation
distance makes the prediction less effective. Ideally this might be accomplished by a solar sail type of mission.