Back in 1996, when NASA solar physicist David Hathaway and others projected that the approaching peak in sunspot activity would be among the most severe in recent times, space weather forecasting was in its infancy. So we can probably excuse the predictions for being a little off the mark.

By applying decades-old data to current models (a well-known process in meteorology circles called hindcasting) Hathaway let history tell him which ones work best in given situations.

Two models produced by other researchers, he learned, are best at predicting the maximum years in advance. Another method, which he has developed, is more accurate at prediction in the months leading up to the maximum.

So Hathaway decided to combine the methods.

The result is a supermodel that can be fine-tuned on the fly as the solar cycle progresses. He explained that he's slowly switching over to giving more weight to his model, which evaluates current activity, and less weight to two models that made their best predictions at the last minimum, back in 1996. Right now the mix is 50-50, but will gradually shift to 60-40.

Hathaway expects the maximum to arrive in June, a prediction that hasn't changed for about four months. High activity will continue thereafter for two or three years, he said.

How strong will it be? Probably not as strong as reports have suggested over the past four years. The official expectation had been a peak of 160 sunspots, a figure that represents a daily average for one entire month, and that is roughly equal to the 1989 maximum. Hathaway now figures the peak will be about 136.

"The current prediction is that this is going to be a bigger-than-average cycle, but not as big as the last two," he said.

January has already had several days with more than 200 sunspots, Hathaway says, and the average for the month will likely be about 130. Each day in December averaged just 86, but in November there were 132 per day. The long-term average is about 100. So far, there have been no reports of major disruptions to satellites and power grids.

One of the models Hathaway has incorporated into his suite comes from Richard Thompson of IPS Radio and Space Services, part of the Australian Government Department of Industry, Science and Resources. Thompson's model was found to work better when contrasted with another.

"Essentially Dave has found that my method and one by Joan Feynman are different enough that the combination of the two gives a better result than either one alone," Thompson said. "I like what Dave has done."

Thompson has recently revised his own estimates. In the middle of last year, he called for the peak to occur in April of this year. Now, he tells space.com, he sees the peak coming later than what Hathaway is predicting.

"I would expect it to be late 2000 or early 2001," Thompson said. He added that while the intensity will probably be less than originally projected, he still expects the sunspot average to exceed 140.

Hathaway says we won't even know when the maximum has arrived until months later. This is because researchers will have to first see the downward slope of the statistical bell curve of activity before they can know we've topped out.