Why This Winter Is Different
This fall's El Niño is at record levels by ocean temperature. September's Niño-3.4 reading came in at +2.84°C, the highest monthly value since 1982. (One note for the weather folks: the official ONI is not a record yet.)
A strong El Niño usually means a mild winter across the northern US. The classic example is 1997–98, which went down as the 6th-warmest Northeast winter on record. That's why so much of the seasonal consensus is leaning warm this year.
This year, I'm leaning against that warm consensus. El Niño isn't the only thing steering the pattern, and several of the other big players look very different from 1997. Here's why.
The Drivers Behind My Forecast
- The PDO is firmly negative. The Pacific Decadal Oscillation is a long-lived pattern of ocean temperatures in the North Pacific. It came in at −1.55 in September. That's a big difference from 1997, and it changes how El Niño's influence plays out over North America.
- The QBO is a near match to fall 1957. The Quasi-Biennial Oscillation is the back-and-forth of the winds high above the tropics in the stratosphere. It affects how stable or disrupted the polar vortex gets during winter. This fall's setup is a near match to 1957.
- The polar vortex is running slightly weak. The polar vortex is the ring of strong winds that keeps the coldest air locked up near the North Pole. Right now it's slightly weak, and models are forecasting a wave-2 displacement toward Hudson Bay around Halloween. That means the vortex gets pushed off the pole toward Canada rather than staying centered.
- A different November look in the Pacific. My read of the Euro Weeklies is a November ridge in the eastern Gulf of Alaska. In 1997, that same area had a trough. That's one of the clearest signs this isn't a repeat of 1997.
The analogs I'm using
Analogs are past winters with a similar setup that help show how this one might play out. Because of everything above, for December and February I've moved off my 1997/1982-led blend toward 1957–58, 1972–73, 1965–66 and 1969–70, while keeping 1997–98 in at a lower weight. January stays with my base blend.
Winter Temperature Outlook
Overall, I'm going near normal to slightly cold in the East and colder than normal in the South. That's a big departure from the typical strong El Niño winter. The story is really in how the season evolves month to month.
Month by Month
December 2026
December comes in near to slightly below normal in the East and South, with the Northeast running about −0.7°F. The northern Plains look milder.
January 2027
January is the mild month, running about +2 to +3°F in the Northeast. That's when I expect the polar vortex to recover and the Pacific jet to take over the pattern.
February 2027
February is the core of winter. Here's the key signal: 4 of my 5 cold analogs had a major sudden stratospheric warming (SSW) in January–February, including late January in both 1958 and 1973.
A sudden stratospheric warming is when the air high over the pole heats up rapidly and disrupts the polar vortex. That can send Arctic air spilling south a few weeks later. If that repeats this winter, February runs about −3°F in the Northeast (for reference, February 1958 was −7°F) and 5°F or more below normal across the South and Gulf Coast.
Winter Snowfall Outlook
The El Niño southern storm track is the story this winter.
- Above normal (125–150%+ of normal) from Texas and Oklahoma through Tennessee, Virginia and North Carolina. Keep in mind those are fairly modest normals to begin with.
- Near to above normal in interior New England.
- Near normal (90–110%) along the I-95 corridor. That's my adjustment above the raw analog average.
- Below normal (50–75%) across the northern Plains, Upper Midwest and Great Lakes.
- 75–90% of normal in the northern Rockies.
February is the wild card. February 1958 buried the Northeast, but most of my analog Februaries were below normal there.
What Could Bust This Forecast
Two things could flip this to a milder, 1997-style winter:
- The Gulf of Alaska trough sticks around into November instead of the ridge I'm expecting.
- The polar vortex recovers with no late-January warming. For context, about 3 in 4 westerly-QBO winters since 1958 had no January–February sudden warming.
If either of those happens, the warm El Niño consensus wins out.
Bottom Line
The record-warm El Niño is pointing toward a mild winter, but the negative PDO, the 1957-like QBO, a slightly weak polar vortex, and a different November Pacific pattern all argue against a 1997 repeat. I'm expecting a near-normal to slightly below normal December, a mild January, and a February that could turn into the core of winter if a sudden stratospheric warming shows up. Snow-wise, the southern storm track is the big story, and I-95 sits near normal.
I'll keep tracking the November pattern and the polar vortex closely and post updates here and on X. You can follow along at @weatherwilly.
