Predicting the solar harvest

The New Scientist has the following report.
The state of the Sun's atmosphere has been predicted with unprecedented accuracy five days in advance, using some of the world's fastest computers. The simulation lays the foundation for better forecasts of hazardous magnetic storms around Earth.

Clouds of gas that escape from the Sun in events called coronal mass ejections (CMEs) can fry satellites and knock out power grids on Earth. The storms also pose a radiation hazard to astronauts.

But predicting these storms is difficult – not every CME that heads our way triggers a storm. The key factor appears to be the alignment of the cloud's magnetic field. If it is lined up with Earth's magnetic field, the local space weather is likely to be calm, while if it is lined up the opposite way, powerful magnetic storms can result. Now, researchers led by Zoran Mikic of the company Science Applications International in San Diego, California, US, have devised a sophisticated computer model based on observations of magnetic activity on the Sun's surface, or photosphere. This activity shapes the Sun's wispy outer atmosphere, or corona, where the eruptions of gas originate.

They improved on simpler versions of their model by including a more accurate simulation of how energy flows through the corona. The new model worked well enough to successfully predict the shape of the entire corona when it became visible in the solar eclipse of 29 March 2006. Normally, the Sun's surface is too bright to directly observe the corona.

While this research is mostly relevant to the high end applications we can also use such data to fine tune solar collectors of different types.