The Spread Maps

Menu

Request an invite

Where the models disagree today

Sunday's widest spread is in Phoenix, where seven models put the high anywhere from 24° to 31° in two camps; every city's spread is on the map, and the day's written read follows it.

How this page is built

Every number here is computed from The Spread's own archive of what each model actually published, run by run. The written discussion narrates those numbers and is not allowed to add any — no fronts, no systems, no forecast. It describes what the models are doing, which is what this archive measures.

Sample is 26 archived US cities, so "everywhere" means those cities. The map draws markers only: interpolating between them would show a pattern nobody measured.

26 archived cities · daily high temperature · day 3 (2026-10-18) · guidance as of 2026-10-15T11:50Z

Where they're fighting

One marker per archived city, labelled with its coolest and warmest model forecast for day 3; bigger and deeper in colour is a wider spread, and a dashed ring means three or more camps.

Reading the map

Each marker is one archived city, placed geographically. The label gives the coolest and warmest model forecast for day 3 — so a wide pair is a city the models disagree about — and marker size and colour are the spread between them. The coastline, lakes and state lines are there to place them; no value is drawn between the cities, because we know these 26 points and not the country between them.

A dashed ring means the models fractured into three or more distinct camps there — not just a wide range, but several competing answers at once. Most cities carry two camps at this lead; these are the ones telling more stories than that.

On a phone the map is drawn at the screen's width and names only the five widest; the list under it has every city.

Seattle 12–15° Minneapolis 16–20° Boston 9–16° Boston Buffalo 13–15° New York 12–16° New York Chicago 14–17° Philadelphia 12–16° San Francisco 16–23° San Francisco Washington 13–17° Denver 20–23° St. Louis 16–20° St. Louis Los Angeles 29–32° Charlotte 14–17° Oklahoma City 15–18° San Diego 28–32° Phoenix 24–31° Phoenix Atlanta 17–19° Dallas 18–22° New Orleans 20–22° Houston 21–24° San Antonio 20–23° Orlando 22–27° Tampa 24–28° Miami 26–29°

The widest spreads

Moved since yesterday: New York, New Orleans

  1. Phoenix24–31°7° spreadtwo camps
  2. San Francisco16–23°7° spreadtwo camps
  3. Boston9–16°7° spreadtwo camps
  4. Orlando22–27°5° spreadtwo camps
  5. St. Louis16–20°4° spreadtwo camps
  6. New York12–16°4° spreadmedian moved 16° → 14° since yesterday
  7. Tampa24–28°4° spread
  8. Salt Lake City18–22°4° spread
All 26 cities
  1. Minneapolis16–20°4° spread
  2. Dallas18–22°4° spread
  3. Philadelphia12–16°4° spread
  4. San Diego28–32°4° spread
  5. Washington13–17°4° spread
  6. Oklahoma City15–18°3° spreadtwo camps
  7. Los Angeles29–32°3° spread
  8. Detroit14–17°3° spread
  9. Chicago14–17°3° spread
  10. Charlotte14–17°3° spread
  11. San Antonio20–23°3° spread
  12. Houston21–24°3° spread
  13. Denver20–23°3° spread
  14. Seattle12–15°3° spread
  15. Miami26–29°3° spread
  16. Atlanta17–19°2° spread
  17. New Orleans20–22°2° spreadmedian moved 23° → 21° since yesterday
  18. Buffalo13–15°2° spread

These are day 3. For today's spread in each of these cities, see every location we archive.

Today's discussion

Phoenix is where the forecast models disagree most about Sunday, three days out. Their highs run from 74.5 to 87.1 degrees, and the models split into two separate camps rather than scattering around one answer. San Francisco is close behind, at 60.8 to 73.2, and so is Boston, at 48.8 to 59.9. Both of those cities also show two camps. When models split like this, a single forecast number for that city is closer to a coin toss between two outcomes than a firm call. In Phoenix, the American GFS model has had the smallest errors at three days out.

The map data hint at what lies behind two of these splits. In Boston, the European ECMWF model gives the coolest high, 49, and NOAA's AI-based AIGFS gives the warmest, 60. ECMWF puts its area of high pressure, at 1035 hPa, about 40 miles northeast of Boston near Salem, Massachusetts. AIGFS puts its version 370 miles southwest, near Annapolis, Maryland. The cooler ECMWF high appears tied to the high sitting so close to the city. In San Francisco, GFS's 62 is cooler than the Canadian GEM model's 67. That looks tied to an area of low pressure GFS draws 240 miles north, near Mt. Shasta, California, which GEM does not draw at all.

Elsewhere, the models mostly agree. Of the 26 cities tracked, 21 have ranges of about 8 degrees or less. Buffalo, Seattle and New Orleans are tightest, at about 3 degrees. Over the past day, the middle forecast for Sunday dropped 3.2 degrees in New York, to 56.9, and 3.1 degrees in New Orleans, to 70.4. The way the models grouped shifted together across Minneapolis, Chicago, Detroit, Buffalo, Philadelphia, New York and Boston, and separately across San Antonio, Dallas and Houston.

On rain, at least one model has a tenth of an inch or more falling over a quarter of the map in the day to Sunday afternoon. All of them agree on only 8 percent of it. Across all the cities, GFS tends to run about a degree warmer than the other models.

Issued 2026-10-15T00:52Z from 26 cities' archived model runs. Written from the numbers on this page — no forecast of its own. Its temperatures are in °F.

Maps that draw every model

Unique maps which illustrate the spread: each one draws multiple models at once. All 24 maps

Which models run warm

How far each model's highs sit from the median of all of them, averaged over every archived city and lead; positive runs warm. This is not accuracy: the median can be wrong too. Warm and Cool count model-days at least 1.7° off the median, out of 52 to 78 each. Consensus products are left out — they are averages of the rows below, not competitors.

Modelvs median WarmCool Model-days
GFS +0.66° 18 3 78
AIGFS +0.07° 6 4 78
ICON -0.03° 1 1 52
UKMO -0.17° 4 10 52
GEM -0.26° 7 13 78
ECMWF -0.27° 11 18 78
AIFS -0.30° 1 7 78