(CNN) After all these years, Bing Crosby still spurs our dreams of a "White Christmas." But a new data analysis suggests some of us are better off leaving that dream of snow on Christmas morning to the holiday classics.
When meteorologists talk about "normal" or "average" weather, they're mainly referring to a data set that takes into account conditions over the past three decades. That so-called normals data was updated this year, and now includes more recent weather — from 1991 to 2020 — which has tended to be milder and less snowy due to the climate crisis.
This week, National Oceanic and Atmospheric Administration (NOAA) scientists used that data to update the chances for at least one inch of snow on the ground on December 25.
So, cross your fingers and use the map below to see if there is still at least a small chance of building Frosty on Christmas Day.
The highest chance of snow on the ground on Christmas Day is about where you'd expect; parts of North Dakota, Minnesota and Wisconsin have chances higher than 90%, along with upstate New York and interior New England.
The cold, mountainous areas of the West also have a high chance of snow on December 25.
But what isn't clear from the map above is where these probabilities have changed. CNN crunched the numbers for more than 2,000 locations that have ever had a chance of a white Christmas in NOAA's normals data.
An incredible 64% of those locations have a decreased chance of a white Christmas in the new normals data compared to the previous 30-year average (which took into account weather from 1981 to 2010).
Only 31% of the stations increased their probability, leaving 4% of the locations staying about the same.
"If you're tempted to compare the two maps for signs of the influence of long-term warming, keep a few things in mind," writes Rebecca Lindsey from NOAA. "From one 30-year normals period to the next, two decades of the data are the same; only one decade out of three is new."
This means the change from one normal to the next can be relatively small since you only change a third of the data. "It also means that at some locations, changes could be the result of natural decadal variability," Lindsey says.
For instance, take New York City in the first half of the 20th century, where you can see this decadal variability at play. Around 1940, when Irving Berlin originally wrote the song "White Christmas," it had been a decade since there had been snow on Christmas in New York City, where he lived.
The city had seen an impressive run of white Christmases from 1911 to 1920 and then again at the end of the '40s. After a lull, the 1960s produced another excellent decade for snow on Christmas in Central Park.
Then the pattern has shifted to a much less frequent cadence. And there's only been four white Christmases in New York City since 1970. Since 2009, the city of the "Miracle on 34th Street" has had no snow on the ground for Santa's landing.
You can't construct a similar timeline for every location. In many places, weather data is extremely sparse prior to the 1980s. But the trend seen here in this new set of white Christmas probabilities echoes the changes scientists have seen across the US.
"The spatial fingerprints of climate change that we'd expect based on our modeling studies are visible in national patterns of change in temperature and precipitation Normals," Lindsey writes. "So, it's not surprising that there are some subtle differences between the 1981-2020 version of the white Christmas map and the 1991-2020 version that are consistent with the reality of long-term warming."
Parts of the South, for example, used to have at least a small chance of seeing snow on Christmas Day. But that chance has faded in the latest data. It makes sense when you look at how average December temperatures have warmed across the Southeast.
The connection to how temperatures have changed is strong across the US. The only area that showed cooler average temperatures in the new normals data was the Northern Plains, where white Christmas chances increased as well.
It may take another decade or two of record-keeping to know exactly how much the climate crisis is impacting your white Christmas dreams. But don't go full Scrooge — there are always outliers.
"While the map shows the historical probability that at least 1 inch of snow will be observed on December 25, the actual conditions in any year may vary widely from these because the weather patterns present will determine the snow on the ground or snowfall on Christmas day," Lindsey optimistically adds. "These probabilities are useful as a guide only to show where snow on the ground is more likely."
Check your chance of snow the next 10 days here
And if you are still wondering — like the Haynes sisters in a "White Christmas" — "What is a Christmas with no snow?" Take it from this writer who lives in the South and just saw his dreams smashed with the new probabilities: A Christmas with no snow is still Christmas!
In the words of Clark Griswold: "We're gonna press on, and we're gonna have the hap- hap- happiest Christmas since Bing Crosby tap-danced with Danny [bleepin'] Kaye."