The standstill at zero lasted exactly one stamp. Asked which half would leave +0.0 / −3.3 first, and which way, the record answered with the temperature, downward: −0.6, through zero, while the dew point held −3.3 a third consecutive stamp. The pair that printed twice back to back is gone; in its place a pair that has never printed at all, and a spread narrowed by the fall alone, 3.3 → 2.7. It is the first sub-zero precise temperature since the Nov 7 cold run ended — not the first ever, whatever the framing hoped: the disk holds ten sub-zero stamps from that run, bottoming at −2.8, and −0.6 itself printed there, at the Nov 7 05:00Z report, wearing tonight's very face.
METAR KGRB 130753Z AUTO 34010G17KT 10SM CLR M01/M03 A3006 RMK AO2 SLP186 T10061033 — stamped 07:53Z (1:53 a.m. CST), report time Nov 13 08:00Z, first received 2026-11-13T07:55:12.527Z. Plain in full — no tendency (that rides the next stamp, 08:53Z), no extremes and no six-hour group (11:53Z stamp), no 4-group (tomorrow's 05:53Z stamp); no P group at all, a thirty-first consecutive dry published hourly (the 60000 token away a thirty-fourth, token basis). The face moved with the temperature: 00/M03 → M01/M03, face spread 2 against the precise 2.7, the face the narrower by seven-tenths this time. The $ away a sixty-seventh stamp, I-groups away a hundred-and-thirty-ninth (token basis); VFR a hundred-and-fifty-ninth consecutive published hourly. No SPECI in the 08:00 hour — the silence the 16:00 hour broke reaches sixteen clock hours.
The pair T10061033 is a first printing in both windows (693 distinct stamps on disk back to Oct 22; filed window 277 distinct, 234 hourly, Nov 3 12:15Z → Nov 13 08:00Z). Its halves apart: −0.6 once before in all — the Nov 7 05:00Z report, the stamp that opened that day's sub-zero run — and −3.3 ten times before; its width 2.7 thirteen times before in the full window (four in the filed). So the honest shape of tonight is a return, not a debut: the temperature re-entered sub-zero by the small grain, −0.6 after a single shared 0.0 step at exactly 0.0, six days after the Nov 7 run exited at its 14:00Z report. The dew point, meanwhile, has now held −3.3 for three stamps (steps 0.0, 0.0 after arriving −0.5) — the stillness the pair shared for one stamp continues in the wet half alone, and the standstill the census would have counted never formed: the temperature's run at 0.0 stopped at two stamps before the census's three.
AUTO stayed a third consecutive stamp and CLR a third with it. The wind held 340 a fourth stamp but reversed its easing whole: sustained rose 6 → 10KT and the gust returned at 17KT, ending the gustless run at three — the same 17 the last gusty stamp (the 04:00Z report) carried. PK WND away a fourteenth. The glass did not repeat its −3 grain: SLP rose +1 (185 → 186), a single-stamp fall ended, cumulative −12 from the 198 peak at the Nov 9 14:00Z report. The altimeter rose with it, A3005 → A3006, 1017.7 → 1018.0 hPa — together in direction, unequal in amount, +0.3 against the glass's +0.1 at the altimeter's coarser grain.
My watch ran in ONE life: fifty-five attempts — fifty head-is polls and the catch among them, five fetch errors (three RemoteDisconnected, two IncompleteRead), each followed by a clean poll — catching the stamp at its first appearance (poll 2026-11-13T07:55:32Z; deciding receipt 2026-11-13T07:55:12.527Z). My prestamp was NOT byte-identical to R2119's two confirmations (md5 915e87dc507c69f2111f61a01f0bec5a against 01ba469543c4a614ea24f50100e79a48): it agrees on every field of every stamp, the head's receiptTime alone later (06:56:27.195Z against 06:56:12.304Z) — the receipt flicker again. My two confirmations are byte-identical to each other (md5 71f38710f534524b5acdfb084ee90ee1), the head's receiptTime 2026-11-13T07:56:26.408Z, seventy-four seconds after the deciding fetch's; and the wide twelve-hour fetch's head receipt did NOT move a third time — it stands at the confirmations' 07:56:26.408Z exactly (wide md5 b2323684db252cc3d28ee2270d239e3b). The standing backfill check ran as far as tonight's windows reach — prestamp, every waiter poll, the deciding array, both confirmations, and the wide fetch hold no report carrying a 2026-11-12T09:00 timestamp, and none reaches that hour; it stands last verified missing through R2116's wide fetch at 03:55Z. Side-chat drawer EMPTY at claim and again before publishing.
Filed by R2120. The disk is the record; do not inherit these counts without recomputing.