Precipitation is chaotic


A team of McGill researchers have refuted an age-old assumption concerning the randomness of precipitation. Scientists have long thought that rain and snowflakes fall in a uniformly random pattern that evens out over large areas – a mistaken hypothesis that is at the heart of modern weather predictions. When an area is hit by a storm, precipitation is considered the same throughout the region.

"Obviously this assumption is inaccurate," says Shaun Lovejoy, a professor of Physics at McGill University and principal author of the research. “It’s easy to see alternations between light and heavy precipitation during a storm.” The rain component of Lovejoy’s research, published in Physical Review, indicates that smooth, constant precipitation does not exist at any observed scale. Instead, rain falls in a pattern of eddies and whirlpools that are known as fractals by a branch of mathematics known as chaos theory.

The study is the first to definitively illustrate that rain is hierarchically coupled to the wind - a new understanding that could one day improve our hourly weather forecasts. Lovejoy anticipates a similar pattern in the snow data currently under analysis.

“That rain and snow might be wind driven is perhaps as obvious as the fact the Earth rotates around the Sun,” says Lovejoy. “Yet this has never been quantitatively documented - particularly at the scale of individual drops or flakes” Lovejoy conducted his landmark experiment on the roof of McGill’s Rutherford Physics Building in Montreal. His research team created three-dimensional models of actual cubes of falling rain using stereo photography; a process where an identical scene is photographed simultaneously from different angles. In order to turn the photos into a 3-D model, every rain drop was unambiguously identified and catalogued (at up to 100,000 particles per picture). Lovejoy says the task was simple for the large, infrequent rain drops.

The difficulty came in identifying the thousands of smaller drops – many of which were less than a millimetre in size. It took Lovejoy’s team seven years of research to refute the uniformly random raindrop theory. Information: lovejoy [at] physics [dot] mcgill [dot] ca (Shaun Lovejoy), 514-398-6537.