Weekend weather really is different.

Analyses of more than 40 years of weather data from around the world reveal that in some regions, the difference between daily high and low temperatures on weekend days varies significantly from the same difference measured on weekdays. Because the weekly variation doesn't line up with any natural cycles, the researchers contend the only explanation for the disparity is human activity.

For part of the new study, the scientists compared the diurnal temperature range (DTR)--the difference between the daytime maximum and nighttime minimum--at 660 selected weather stations in the continental United States. At each of more than 230 of those sites, the average DTR for Saturday, Sunday, and Monday varied from the average DTR for Wednesday, Thursday, and Friday by several tenths of a Celsius degree, says Piers M. de F. Forster, a climatologist at the National Oceanic and Atmospheric Administration in Boulder, Colo. Storm systems moving across an area can cause shortterm fluctuations in DTR, but those variations shouldn't consistently fall on certain days of the week, says Forster.

He and his colleague Susan Solomon of the University of Reading in England describe their findings in the Sept. 30 Proceedings of the National Academy of Sciences.

The cause of the effect isn't clear, but geographical patterns may provide a clue. While sites in the southwestern United States typically recorded a broader DTR on the weekend, those in the Midwest--a couple of days downwind--experienced a wider DTR on weekdays. Windborne pollutants from southern California, for example, first could be affecting weather close to home and later influencing midwestern weather, says Forster.--S.P.