Insights & News

Spring Snows in Europe & The USA: A Dwindling Story

12th March 2021

Across much of Europe and the USA, lying snow is proving increasingly scarce during the spring season. Losses are largest where the most days used to be seen: in the northern regions. Read on to discover just how big those reductions are and what this means from a risk perspective…

Spring Back & Forth

Ah, the spring season. A time of astronomical positivity: Lengthening days, increasing strength of received sunlight. Temperatures responding in due course.

Except, it’s not usually that simple. Short-term weather patterns dominate our day-to-day experiences and that includes how warm it is.

In many parts of the world, it’s not uncommon for wintry conditions to hang on well past the meteorological (Feb or Nov) or astrological (spring equinox) ‘end’ date. Even when it does depart, it has a habit of biting back at least once.

This all means that snow is far from a rarity, especially earlier in the season and at the higher latitudes (further from the equator). As you’ll see later, the northeast of both the USA and Europe can see dozens of days with lying snow.

Forth More Strongly

In recent decades, though, we’ve observed a substantial shift in the balance of power.

The seasonal advance of conditions too warm for snow has trended stronger and faster. That’s the inescapable conclusion drawn from analysis of cleaned, quality-assured data from 220 weather stations in the USA and Europe. The results of which, are presented in the remainder of this blog piece.


Europe: A Downhill Slide – Especially in the North

This study examines the typical number of days in March and April with at least 1 cm of lying snow.

Below, you can see how this compares for Europe between two periods: 1973-1999 and 2000-2020.

If you’re wondering why the first period starts in 1973, it’s because that’s the year that a vast network of stations known as SYNOP was activated, greatly increasing the number of weather stations delivering good quality reports.

Click on an image to view in full resolution.

The main thing to focus on here is the coverage of blue (10-20 days), white (20-30) and pink (over 30) shaded circles on the map.

See how it shrinks generally north-eastward when comparing 2000-2020 with 1973-1999. You can see just how large the changes are on the third map. Much of Central Europe has tended to see 5-10 fewer days with at least 1 cm lying snow. Some have lost more than 20.

In Scandinavia, reductions of 20 or more days are widespread in Sweden, 10-20 days in Norway and Finland. We’re taking around a half to two-thirds of a month’s worth!

This descent toward scarce snow cover is also clearly seen when statistically analysing on a by-decade basis:

On these graphs, the length of the yellow bars indicates the spread of the regional station averages.

Clearly, northern Europe sees a lot of variation between the years. Despite this, clear trends can be seen over time.

The least clear is that for Northwest Europe, owing to a relatively low-snow run of years in the 1970s. From the 1980s, we then see a downward, accelerating trend. Even so, taking this region in isolation, you could argue that natural cycles may be driving the changes.

That falters when looking at Northeast Europe, however. This region has seen an astounding decline in the median (red bars). The 1970s through 90s tended to see 40-50 days with at least 1 cm lying snow. In the 2000s this cascaded to 15, with a further decline to just 8 in the 2010s. An 80 to 84% loss!

Climate Change Connections

This is tied to a warming climate in two ways. The first is reduced snowfall in winter (detailed in my early December blog). The other is a shift toward rain, rather than snow, in the spring season.

Generally, this means fewer late accumulations of snow disrupting events and construction projects. In the north, this is counterbalanced by faster melt of the winter snowpack, increasing the risk of disruptive flooding.

In the southern regions of Europe, lying snow is much less frequent. This makes trends difficult to spot – yet we can still see some suggestion in the spreads. In Southeast Europe, the upper quartiles (top of the yellow bars) have been gradually decreasing, with the 2010s 35% lower than the 1980s.

The Contiguous US: More Mixed but Still Generally Less Snowy

Same process, new area. The contiguous United States also has an interesting story to tell.

Here, changes are more subtle compared to Europe, as there’s a narrower area in which temperature trends have much impact on snowfall*.

Even so, we can still see substantial changes, mainly in the north. Specifically, north of a line from Sacramento to Pittsburgh, most locations have seen a drop of at least 10 days in 2000-2020 compared to 1973-1999.

Numerous places have lost more than 20 days with at least 1 cm of lying snow. Again, we’re talking at least a third of a month’s worth! Notably, this includes the Sierra Nevada, a vital warm season water source for parts of the west. We’re talking an increased risk of water shortages in Las Vegas, for example.

Like with Europe, the decadal, regional breakdown reveals some regional variation on the theme:

In the northeast, the 1970s were less snowy than the following two decades. A less prominent imitation of what we saw for Northwest Europe.

Atlantic Relation

This pattern could be largely due to changes in the level of sulphates in the atmosphere, which relates to human activities and hit a peak in the 1980s-early 1990s. That may have temporarily cooled the North Atlantic, increasing springtime snowfall in adjacent regions (tied into the so-called Atlantic Multidecadal Oscillation, which may not be an oscillation at all).

In both the northwest and southeast USA regions, we can see a strong declining trend in the upper quartiles. Meanwhile, the medians haven’t changed much. That tells us that the change is largely within the snowiest spring seasons. In other words, a cold spring used to mean a lot more lying snow than it does now.

From a risk perspective, the ‘ceiling’ to disruptive impacts has lowered substantially. So, for example, the worst springs for construction project downtime aren’t as bad as they once were.

Looking to the southeast region, what we see for the 2010s is not an artefact. Not one station in any year that decade has reported at least 1 cm of lying snow. While we can’t be sure this hasn’t happened prior to the 1980s (not enough quality data), it’s still a startling observation.

The Future is Green?

In the past 47 years, many locations in the northern parts of Europe and the USA have lost 20 or more days with at least 1 cm lying snow, having typically seen 20-30 days during the 2010s. A linear extrapolation – steady trend – from there suggests most of those reach 0 days by the 2070s.

Imagine that – little snow cover to be found in the Arctic reaches of Finland, or around the Great Lakes, in March!

Even if the trends slow down, there are a lot of locations facing a loss of most or all springtime snow cover by the mid-21st Century.

That’s a big deal for human water management (e.g. Sierra Nevada), wildlife habitats (hunting, gathering and hibernation cycles) and the seasonal advance of warm, then hot weather from the south (snow melt takes up a lot of heat energy, slowing that advance), to name a few.

James Peacock MSc
Head Meteorologist at MetSwift


* This has much to do with the USA being both located closer to the equator and having more landmass to the north. This means that in spring, the south tends to be hotter, yet the north colder, compared to Europe. So, the temperature changes faster as you head north or south. There’s less distance between where it’s already too warm and where it’s still cold enough despite the climate warming so far.

Speak to one of our experts and discover how our analytic platform can revolutionise your weather & NatCat insight. Request a meeting, a chat & demo, or a free trial. Developed via consultation & collaboration.