Each raindrop may not be unique like a snowflake but it can be different enough to tell where it came from, according to U.S. Department of Agriculture scientist, Harry B. Pionke. "Differences in the weights of oxygen atoms in water molecules act as fingerprints for raindrops. We are tracing the paths of raindrops through soil to groundwater and streams by comparing oxygen atoms in water samples," he said. "In the past, we've added tracers such as bromide salts or dyes to water," he said. "This works fine on a plot or small-scale field, but you need to add many tons and run long-term experiments to follow water through an entire landscape or watershed," said Pionke, a soil scientist with the Agricultural Research Service. But, he added, "when you want to look at that much land, it's far more practical and less time-consuming and costly to rely on something that's already in the water."

He said the fingerprinting is based on the fact that about 99.8 percent of the oxygen in nature has a weight known as O16. A heavier version -O18- comprises another 0.1 percent, with the rest being atoms of other weights. He said the ratio of 016 to 018 varies in rain from storm to storm, sometimes within a storm and from season to season. Oxygen atoms increase the accuracy of tracking rainwater because foreign tracer chemicals can escape the water by binding to or degrading in soil, Pionke said. This distorts results. Natural oxygen used as a tracer is part of the water molecule, he said, so it will go wherever water goes. "That's important because we must first understand water movement in order to predict chemical movement," he said. "Without knowing which fields are contaminating groundwater or streams with fertilizers or pesticides, how can you identify problem areas and recommend solutions to farmers?" asked Pionke.

Pionke, research leader at the Northeast Watershed Research Center in University Park, Pa., said he and colleagues have been successfully using oxygen to trace water samples back to recent rainstorms. He said the research, designed to identify sites causing problems, has received funding under a national water quality initiative begun in 1989. Samples are taken of water that drains off fields, soil water below crop roots, shallow and deep groundwater and spring and stream water. They pass these samples through a mass spectrometer to separate the oxygen atoms by weight. Then they compare the ratios to those in present and past rainfalls. Fingerprinting raindrops will help find the most active land sources of surface runoff to streams and of recharge to groundwater. While the "hot spots" may comprise as little as 10 percent of a watershed, he said, they can contribute most of the runoff. "Some areas within the watershed are much more important contributors than others," he said. "If you control pollution in these active areas, you have the best chance of improving the water quality over much of the watershed, at a fraction of the cost of treating the whole watershed. Information gained from this research will be used to develop and verify computer models that predict water movement. It will help us choose the most cost-effective prediction techniques and give us the confidence to use them," he said. The computer models and accompanying practical recommendations to come from the national program will help farmers comply with state pollution management programs authorized by the federal Clean Water Act of 1987.

Harry B. Pionke, can be contacted at (814) 865-2048.