You may have noticed that we’re getting less snow in winter as the climate warms. But how much? The latest estimate is that the maximum amount of snow that collects on the ground each year in non-alpine areas of North America has been decreasing by 46 billion tonnes per decade since 1980, a new study reports. But interestingly, Eurasia’s annual snow has held steady.

The total amount of snow that falls, or the "snow mass" is important for all kinds of planning because it’s the amount of water released in the spring when the snow melts, says Chris Derksen, a research scientist at Environment and Climate Change Canada who co-authored the study published Wednesday in Nature.

"So this is the water that’s available for hydropower generation, for human consumption," he said. "It’s also the snow melt that can provide a hazard through spring floods."

Because of that, knowing snow mass is important for managing reservoir levels and mitigating floods.

Previous measurements of snow mass have relied heavily on satellite data. But while it’s easy to tell how much surface is covered by snow using satellites, measuring the depth — and therefore the mass — over large areas has been trickier, Derksen said.

"It’s been kind of a holy grail type of thing that we’ve really had a hard time nailing down over the years."

The study, led by researchers at the Finnish Meteorological Institute, combined satellite data with physical measurements by scientists on the ground to get what they believe to be the best measurement yet of snow mass in the northern hemisphere north of 40 N latitude (the latitude of Philadelphia or Madrid).

Satellites are biased

Snow scatters microwave radiation, so satellites can measure its coverage and depth with microwave signals — even in inaccessible areas like the Arctic that are hard to measure on the ground, Derksen said. But when the snow is deep, satellites tend to underestimate it.

By using measurements made by people digging and measuring on the ground, the researchers corrected for that bias.

"The value of those measurements together is much more significant than if we use just the satellite data alone or just the ground measurements alone," Derksen said. He noted that sometimes there is pressure in government to invest in either satellites or the ground measurement network. "What this shows is you need both."

Measurements taken in March, the month with the most snow on the ground, showed an average peak of 3,062 billion tonnes across the northern hemisphere between 1980 and 2018.

But in North America, the amount fell by about 46 billion tonnes per decade.

While that doesn’t sound huge compared to the total snow, Derksen said, "it still gives us pause for concern over time because this is such an important water resource."

Maps in the study show a wide variation across the hemisphere, with big declines in southern Ontario, Yukon and Alaska and around Hudson Bay.

Despite declines in Europe, Eurasia overall hasn’t seen an overall reduction in snowfall over the past four decades, partly because of heavy snowfalls in eastern Siberia.

Derksen said that while a warming climate means the snow season begins later and ends earlier, that doesn’t necessarily mean less snow. That’s because warmer temperatures allow for more moisture in the air, and therefore more precipitation, even in winter when it falls as snow. He predicts in the future, declines in Southern Canada may be offset by more snow in the north.

New satellite needed

While the study includes most flatter areas, such as boreal forest and prairies, it doesn’t include mountains because the satellites don’t have a very good resolution — the smallest area they can measure is about 25 kilometres by 25 kilometres — and that’s not good enough for terrain that’s very variable. While that’s only about 5 to 10 per cent of the land on the map, Derksen said some estimates suggest it may represent closer to 30 to 40 per cent of the snow mass.

Derksen called that a "critical gap" that researchers hope to address with a new satellite that can do more detailed measurements. He said they’re working with the Canadian Space Agency and industry to develop such a satellite. "Our hope is that we can get some traction with this mission."

Robert Way is an assistant professor of geography at Queen’s University who uses snow data and conducts snow measurements for his research on permafrost, but was not involved in the study.

He called the snow mass measurements "a big improvement" over previous measurements of snow mass across the hemisphere. He said the peak snow mass matters a lot for those who rely on the meltwater, such as hydro companies, and the information will be useful for understanding how snow is changing on both a wider and regional scale.

"They’ve shown by having more observations on the ground, they can improve this data product quite a bit," he added, "which highlights the need to continue to record measurements of these critical variables on the ground."

Way said reliable snow mass data can be used not just for water resource management, but to validate climate models, and work should continue to improve the data.

This study, he said, "is a big improvement over previous efforts. But it’s still incremental on the road to being able to better understand snow and how it’s changing in response to climate change."