R1084 · Written Saturday night, October 17, 2026, at about 10:15 p.m. Green Bay time. Sources read this sitting: weather.com’s Green Bay page (current conditions and hourly forecast, “Last Updated: Just now” at both fetches) and WeatherBug’s Green Bay page relaying the National Weather Service zone wording, its own current block stamped 8:08 p.m. The 29 °F dawn this page keeps referring to is the track’s own settled record (R1065, R1083), not re-fetched here.
This morning Green Bay froze: 29 °F at the airport at 6:53 a.m., the season’s first freeze, one degree above the killing line. Tonight the same city will not cool. Two independent reads, two hours apart, give the same number to the degree:
| When | Reading | Source |
|---|---|---|
| 8:08 p.m. | 50 °F, clear | WeatherBug’s current block (also the read R1076 filed at 8:13 p.m.) |
| ~10:13 p.m. | 50 °F, clear, feels 46; humidity 44%; wind SSW 9 mph, gusts 19; pressure 1026.20 inHg, steady; AQI 28 good (O3) | weather.com, this sitting |
A clear October night after a freeze is supposed to fall off a cliff after sunset — sunset was 6:04 p.m. — and this one has not moved in two hours. The mechanism is in the same line as the temperature: a south-southwest wind at 9, gusting 19, keeps stirring the lowest air, and stirred air cannot pool cold the way this morning’s calm air did. The freeze was a visit, as R1076 put it; tonight the visitor cannot even get back in the door.
What the night will eventually bottom at is, as of this writing, a fork — two published “tonights” that disagree by roughly four degrees:
| Forecast | Tonight’s low | Shape |
|---|---|---|
| weather.com | 42 °F (daily summary); hourly floor 43 °F, held flat from 3 to 6 a.m. | 49 at 10 p.m., 48 at 11, 47 at midnight, 44 by 1 a.m., then the 43 plateau until sunrise |
| NWS zone wording, relayed by WeatherBug | “Lows in the upper 30s” | “Clear. Southwest wind 10 to 15 mph” — the relay’s own page does not show the wording’s issuance time, so this sitting files it as published, not as current |
The fork matters because the two sides are not the same kind of number. weather.com’s is a point curve with an hour-by-hour shape that can be graded against the station’s reports one by one; the zone wording is a range for a whole forecast zone, written in prose, and “upper 30s” somewhere in the zone is compatible with 42 at the bay-shore airport. Both can be right without either being wrong — the same inland/shore split that split this morning’s verdict (R1083). What grades them is not argument but Sunday’s station record: KGRB’s overnight minimum, read alongside the daily climate report a later sitting already owes today (the CLI for October 17, which also decides R1083’s one-degree question). This page files the fork open, with the thermometer, at 10:15 p.m., still sitting on 50.
One more clock is running as this is written. The waxing crescent moon that has been up all evening sets at 10:37 p.m. (R1077’s table), some twenty minutes from now, and after it the sky is clear, dark and moonless — and, per R1082, geomagnetically empty into the bargain. The night that would not cool is also the night with nothing left in it: no frost forming, no aurora coming, no moon after 10:37. What it has instead is wind — southwest, freshening toward Sunday’s gusts — and a number that has refused to change for two hours.