Yet Another Big Solar Flare

philthedrill writes "CNN is reporting that the sun has fired another large solar flare towards Earth. This one could arrive as early as Thursday (Oct. 30th) afternoon. (insert end-of-the-world statement here)."

Comments

Pity

[nepheles]

It's a pity that more people don't get to view these phenomena, because of modern light pollution. How many people are losing awe-inspiring sights, such as this and the milky-way?

We don't know squat.

[RobertB-DC]

From the article:
Space weather forecasters say this spate of strong solar flares is not consistent with normal solar behavior. The sun, which follows an 11-year activity cycle, has been quieting down since the last peak in 2000.

Although we humans have been looking at the sun since before we climbed out of the trees (and our moms have been telling us not to even longer), it's almost silly to say that any observation of our local star is "not consistent with normal solar behavior." Just how many of those 11-year cycles have been recorded?

If the ancient Chinese were using pinhole solar viewers to count sunspots for the past 5000 years, that would be one thing. But as has been posted in every Slashdot story on the subject, we have maybe 200 years of scientific data (of varying quality) out of the sun's five billion year history. Even W's pollsters would tell you that sample size is too small.

Statistical anomaly? More like a bad model

[fname]

When I read this line, "I have not seen anything like it in my entire career as a solar physicist. The probability of this happening is so low that it is a statistical anomaly.", I cringed. Either a) This guy thinks his model is great & really believes that this is a 1/1,000,000,000 event, or b) he knows that this indicates a whole in his model, but the reporter ripped the quote out of context to make the story more sensational.

A little statistics primer is in order. In order to quote odds on anything, a statistician needs a model, generally based on existing data. When there is an event that's off the charts, it will usually indicate to the scientist (or engineer) that there is a deficiency in their model (or their process is out of control, for manufacturing types). If I were a solas scientist, this event would indicate to me that the model is not adequate for predicting this sort of thing. Which makes sense, since we probably only have 40 years worth of data; you expect to be thrown a curve-ball every now and then.

So I doubt it really is a statistical anomanly; maybe these solar-flare pairs occur every 50 years or so & that's why we haven't seen it before. But either the reporter needs to better explain the meaning of the quote, or the quotee needs to take a basic course on the limits of statistical probability.