As demand for power grows, utilities are constantly searching for the most efficient ways to generate that power. Cloud physicist Man Kong (Peter) Yau is helping Canadian hydro companies manage their power-generating resources more precisely through improved forecasting of rainfall amounts.
Yau, a professor at McGill University in Montreal, creates computer models based on the physics of weather that can predict how much precipitation will fall in a given region over a 24- to 48-hour period.
Accurate short-term forecasts are critical for utilities like Hydro-Québec because they enable the company to adjust the volume of water flowing into the turbines at its giant dams, which generate electricity not just for the province of Quebec, but also for customers in Ontario and Manitoba.
“There’s a certain optimum flow rate at which the turbines are most efficient,” explains Yau, whose research is supported by both NSERC and Hydro-Québec. “There has to be some knowledge of precipitation in order to control that.”
Yau develops new algorithms to improve the computer models that Environment Canada is already using to make its weather forecasts. Traditionally, the observations that supply the figures that Environment Canada plugs into its meteorological models are obtained by releasing weather balloons 100 kilometres apart into the atmosphere. Yau is trying to improve the initial observations by using radar to produce high-resolution observations of clouds and precipitation, scanned every five minutes.
But changing the initial data entered into the existing computer models causes problems with data assimilation, so Yau is improving the models. He and his students are creating what he calls a “regional ensemble prediction system” with Environment Canada. The system will calculate all the probable amounts of precipitation that will fall on the surface of a particular region during a given time frame.
To forecast the flow of water into reservoirs, Yau also needs to be able to predict a river’s flow. He’s doing that by combining his atmospheric ensemble prediction with a similar model that calculates probable distribution of water through rivers and streams. That kind of system could also help to predict flooding and other extreme weather events.
Yau and the students in his lab are also studying the physical processes behind precipitation systems, such as hurricanes. By understanding how clouds work and how they affect those large-scale weather systems, Yau is better able to predict the amount of short-term precipitation that will fall.
In addition to helping Hydro-Québec, Yau hopes his precipitation modelling will ultimately benefit farmers and large-scale agriculture businesses, as well as air traffic controllers. “These are all different areas that could use this type of improved forecast,” Yau says.
Yau’s interest in detailed forecast models began after high school, when he worked in a meteorological office in Hong Kong, where he grew up. “Quantitative precipitation forecasting is one of the unsolved problems in atmospheric science,” says Yau. “I like solving problems.”