Have you ever wondered why the weather report is so erratic? Sometimes, the guys with the seals (degreed meteorologists), can't get it right, no matter what they do. The past couple of weeks are a good example.

       Not one weather service, including the US weather guys, have gotten their Carolina 6-12 hour forecasts right for over 13 days. Their 4 hour forecasts have been muddy at best. There was one exception - a certain forecaster at a pay weather service got his 4-hours right on the button, but his long range and over 6 hour forecasts were way off. There are very good reasons for all of this and only this guy will admit why.

       I am biased, of course, because this forecaster is a personal friend. For a while, I talked to him every day. I still talk to him from time to time. He is as learned as anybody in the weather field and far more honest than most.

       Most forecasters won't tell you that any forecast over 6 hours is a glorified best guess. This guy will.

       I have actually known meteorologists who will downface you to tell you, you are just ignorant and don't know the theory and formulae behind it all. But they are the biggest fakes of all. And you don't need me to tell you that. Just look at the hair some of these guys have.

       Why can't anybody predict better? Well, it's all because of Mr. Chaos. See, Chaos Theory predicts that when you have too many variables, you just can't make accurate predictions. It's been known for some time in weather predictions. Recently, it was also proven to hold true in numbers theory - what we would call "statistics."

       Simply put, there are too many variables to predict weather accurately more than 4-6 hours in advance. General weather prediction says something like, "The temperature is expected to go down a little and we might see some precipitation." That's fine. What you can't say, 24 hours out, is "It will be 30 degrees and icing," and have it work.

       Think about it. South Carolina is a natural savannah. That means that sometime in the next 40,000 years - give or take 5,000 years - some of the area from the mountains to the sea will become a desert for some amount of time. That's IF EVERYTHING remains the same. That's almost assured NOT to happen. So we may not wind up like Mogadishu. To say anything else is rubbish.

       Global warming is another joke. Nobody knows. Nobody. And even if it will happen, we may not have caused it at all.

       OK, but what about the local, immediate weather? Same thing. The Piedmont has the mountains to change the formulae's results. The Midlands have the wierd winds caused by the super-drying effect of the Piedmont and the Atlanta corridor. The Coast? Lots and lots of variables from many sides. And that's if the info is correct, which it can't be.

       It's not a plot, it's just science. To make the formulae really work you would need fifty times the number of weather stations gathering information every tenth of a second. Then, we could predict weather for maybe 12 hours ahead, accurately. The cost, however, would be prohibitive. But you know me. I have a suggestion.

       Since we have computers that could handle the information right now, why not put up a few satellites that do nothing but read image to info like our space probes do. You would have to rewrite the algorithms but the formulae are already there, so half your work is done. You could literally take readings every hundredth of a second over every square mile of the earth. We're talking terraflops of information, but we have what it takes to do it as soon as we make the modifications.

       But we don't have it yet. And the next time some jerk with a seal tells you what the weather is going to be, just laugh along with the rest of the scientists or give a knowing smile and remember: you can't predict weather accurately more than 4 to 6 hours in advance.

       But maybe, MAYBE, with new discoveries in electro-gravitic and electro-magnetic sciences, we may even be able to forecast as far ahead as two days, maybe.

       But I'd settle for 12 hours.