mSWiM has been extensively validated on a statistical basis using 12 years of Pioneer, Voyager, Ulysses and Cassini data from 3.5 to 10 AU [Zieger and Hansen, 2008].
Main statistical results to keep in mind:
- The prediction efficiency is expected to be highest at the time of apparent opposition, because at this
particular time, the model uses the actual (not rotated) boundary conditions observed at 1 AU that include
possible transient events as well. The apparent opposition is defined here as the time of opposition of Earth
and a given planet or any other body plus the solar wind propagation time from Earth to the body at an average
speed of 500 km/s. See the list of oppositions and apparent oppositions for different planets on the
Optimal Periods page.
- The most reliable predictions are expected within 75 days from apparent opposition. Correlations between predicted
and observed solar wind variables outside of this range are considerably lower but still statistically significant.
Please be extremely cautious when using predictions with large angular separation between Earth and the respective planet.
These predictions may be realistic only under conditions of high recurrence index in the solar wind speed, when the solar
corona is exceptionally stable with hardly any transient events like interplanetary coronal mass ejections (ICMEs).
- The most accurately predicted solar wind variable is the solar wind velocity. The IMF magnitude, density and
the tangential component of IMF (BT) are the next most accurate in the respective order.
The predictions of the temperature must be treated with special caution. Although correlations between predicted and observed temperatures are often relatively high,
the error in the absolute values can be as high as one order of magnitude in some cases. Note that the predictions of the normal component of IMF (BN) are very
poor as shown by the hardly significant correlations between predicted and observed BN. Finally the user must remember that BR cannot be predicted with 1-D MHD
models because of the ∇·B = 0 constraint.
- It is important to keep in mind that the prediction efficiency is highly dependent on the recurrence index of
solar wind speed, which is simply the cross correlation of the observed solar wind speed during
one solar rotation with the solar wind speed in the preceding rotation.
This naturally comes from the model assumption that the solar corona is in steady state on the time scale of half a solar rotation.
Thus the highest prediction efficiency is expected during the late declining phase of the solar cycle when long-lived high-speed
recurrent streams dominate the corona in the ecliptic plane (See Fig. 2)
- As shown by the statistical analysis, the error of shock arrival times is as small as 10-15 hours within 75 days
from apparent opposition in years with high recurrence index.
At solar maximum, however, predicted shocks tend to delay by at least 10 hours,
which implies extra shock acceleration in the real solar wind that is not taken into account in mSWiM.
- Finally, the user must be aware of existing data gaps in the boundary conditions (OMNI input data) that can seriously influence the solar wind predictions. A data coverage of 80 % or higher is essential to ensure reliable solar wind prediction.
Tóth, G. (1996), A General Code for MHD Flows on Parallel Computers:
Versalite Advection Code, Astrophys.Lett.Comm., 34, 245-250.
Zieger and Hansen (2008) Statistical validation of a solar wind propagation
model from 1 to 10 AU, J.Geophys.Res., doi:10.1029/2008JA013046.
Published in Journal of Geophysical Research - Space
Physics. Copyright 2008 American Geophysical Union. Further reproduction or
electronic distribution is not permitted.