"I went into this," says New Yorker Dennis Eskow, "with a prejudice against meteorologists.

"Very often you'll get a forecast for a major storm coming and you'll make your plans around that, only to find that you don't get a drop of anything. It just doesn't materialize."

Eskow, 36, is one of a growing number of amateur weather forecasters. His interest was sparked when a totally erroneous forecast caused him to cancel a star photography expedition.

"Some friends and I were going to go and photograph some binary and triple stars. Nasty weather wasn't forecast for two days. There were six or seven of us and we were hauling heavy telescopes and cameras."

Eskow and his friends wound up being washed out for three days by a sustained storm. "I just thought, 'These guys just don't know what they're doing. Maybe we're all better off being our own weatherman. I want to find out what the problem is.' "

Although convinced that the amateur forecaster can more precisely predict conditions in his own back yard, so to speak, Eskow hastens to add that he is not disparaging professional meteorologists. "I am in awe of them. Nobody but a highly trained person can make a prediction beyond 72 hours. Meteorology is an art and it's a science. I'm impressed with them the way I am with a good doctor.

""Gordon Barnes, for example, is a genius," Eskow says of the ubiquitous TV forecaster. "I call him the Denton Cooley of meteorology. He combines a certain kind of savoir faire about the nature of things with what he knows about the science of meteorology. He's so accurate."

Eskow backs up his belief in amateurs, however, in "Make Your Own Weather Forecasts," an article in the March issue of Popular Mechanics magazine, where he is science editor.

By combining the meteorologist's "barometric reading with the local wind conditions in your area," he claims, "you can use the weather chart (see box above) to get a more accurate forecast of the weather in your own neighborhood for 24 hours."

The forecasting chart came from a number of research sources, several textbooks and interviews with meteorologists around the country. The chart, combined with the barometric and wind data, could result in very accurate--"almost 100 percent"--forecasts in the spring and fall. It would not, Eskow says, always allow you to foresee extreme conditions, such as tornado activity in midwinter or midsummer.

Eskow recommends two inexpensive but "very helpful" tools for the novice forecaster. "An amateur could use the Sager Weather Forecaster ($9.95). It's remarkably accurate for 24-hour forecasts," he says. "If you have that and Pocket Weather Trends' Pocket Weather Forecaster ($3) and a weather chart such as ours, or one you might find in a textbook, that would be excellent equipment for not very much money."

Add $50 to the basic tools, Eskow says, and you can have a "really good wind-direction device that's electronically based. Add an accurate but not very expensive barometer and you'd have a good, worthwhile system."

And if you're really serious about the matter, you could get a home weather-forecasting station for $399 to $1,500.

The expensive, computerized systems, says Eskow, even provide data on degree-days, "the measure of how often and for how long your oil burner had to click on during the winter."

A simple forecasting system could be particularly useful to the commuter: "Say you're a commuter living in Maryland or Virginia. The forecast you're hearing on Washington TV or radio probably pertains strongly to Washington itself, maybe even the downtown area." If you do your own forecasting, says Eskow, you can predict what the weather's going to be at your house, in your neighborhood.

Eskow himself has become an avid forecaster. "I'm quite addicted to it. I beat three local forecasters with next-day predictions in my neighborhood four separate times in a three-week period. Our forecasts agreed the other 17 days, and only once did the professionals disagree in their forecasts."

Almanacs, Eskow claims, are interesting to read, but the Farmer's Almanac, for instance, "really hasn't much use in forecasting."

Folk forecasting, on the other hand, "shouldn't be discarded too easily. The aunt with arthritis, the guy with the arthritic knee are walking barometers. Anybody who gets headaches periodically could test this. Very often the causes for periodic headaches are weather-related."

You don't have an arthritic friend or relative, but do have a cow? You could follow the example of a Texas rancher for do-it-yourself forecasting.

"When his cow's tail is active," says Eskow, "the rancher knows it's going to rain. When its tail is down, the guy knows the weather is going to change for the better."