What does the weather we have been having tell us about the coming winter season?
Autumn is a time of transition. Summer heat persists while winter cold builds, and sharp contrasts can result, like the one in Denver last week: The temperature there dropped from 80 degrees Fahrenheit on Wednesday afternoon to 16 degrees on Thursday morning.
Arctic air pushed into the city behind a cold front trailing a storm — the same storm that caused blizzard conditions in the upper Midwest two days later. That system was followed by the storm that developed into a “bomb cyclone,” lashing the Northeast with high winds and rain on Wednesday and Thursday.
As Denver froze, the same Arctic air mass extended west to the Sierra Nevada, where it drove the downslope “Diablo” winds that blasted Northern California like a giant hair dryer. The hot, dry, gusty conditions there prompted Pacific Gas and Electric to cut power to millions of people, in the hope of avoiding catastrophic wildfires touched off by downed power lines, like those in 2018.
These weather events are predictable in the short term. Coloradans knew days in advance that the mercury would plummet, and accurate forecasts of severe fire weather in California let PG&E cut the power in time to be effective.
But do they tell us anything about what kind of winter will be coming? Probably not.
Individual weather events are not generally predictive of anything over the longer term. Bitter October cold snaps, early blizzards and roaring Diablo winds can occur as readily before a mild winter as a harsh one.
However, we do have forecast models that can tell us — in a general, broad-brush way — whether the upcoming seasonis likely to be different from an average winter, and if so, how. These models factor in what is going on not just in the atmosphere, but also in the ocean, which changes more slowly and thus contains more “memory.”
The seasonal forecasts put out by the National Weather Service’s Climate Prediction Center include maps with broad areas colored to indicate the odds that temperatures or precipitation there will tend to run above (red) or below (blue) those of a normal winter.
Because the uncertainties in the data only permit a very fuzzy view of the future on these time scales, the odds in the forecast tend not to be very high. The center’s national forecast of precipitation for the next three months, for example, shows greater than average chances of wet (or snowy) conditions for the Mid-Atlantic states, Florida and the northern Plains, with Oregon and northernmost California drier than average. But the map does not show any values higher than around 60 percent, meaning that the forecasters are only 60 percent confident that the forecast will prove correct.
That is not much better than a 50-50 coin flip, but even low-confidence forecasts can be valuable, as long as you do not expect them to be better than they are.
Everyone can make one basic seasonal forecast with great confidence, though — that the winter will be colder than the summer. We know this from experience, of course, but scientists also know the essential physical reason behind the change: Sunlight decreases from summer to fall to winter as the Earth’s travel through its orbit swings the Northern Hemisphere to tilt away from the sun. So as we pull out our winter jackets, we are making a prediction of a change in the climate that is caused by what scientists call “radiative forcing.”
You can probably see where I am going with this. Predictions of global warming are inherently similar to the predictions of changing seasons that we all take for granted.
With global warming, the changes in incoming radiation caused by rising concentrations of greenhouse gases occur in the infrared part of the spectrum rather than the visible-light part, and the scale of the change is much smaller from year to year than the swings in the seasonal cycle. But the general idea is about the same, and so are the methods we use to predict its effects. Keep this in mind the next time you hear someone say, “If meteorologists can’t predict the weather more than a week in advance, how can they predict climate change?”
In fact, it is possible to see global warming even in the seasonal forecast. The climate center has issued temperature forecasts for every three-month period between now and a year from now, and all of the maps show only red or white, with no blue shading anywhere in the continental United States. That means all the forecasts are for conditions that are warmer, or at least no cooler, than the average for the last 30 years.
The confidence levels for the temperature forecasts are much higher than those for precipitation — as high as 80 percent in the Southwest for the coming three months. And the fact is, a forecast of higher average temperatures would be a pretty good bet even if it were for one, two, five or 10 years from now. The planet is warmer now than in any previous decade on record, it will continue to get even warmer, and the forecasting models and the scientists who use them know why.
Adam Sobel is a professor and director of the Initiative on Extreme Weather and Climate at Columbia University. He is an atmospheric scientist and author of “Storm Surge,” a book about Superstorm Sandy.