The stamp. METAR KGRB 051653Z 36014KT 10SM BKN012 BKN020 OVC065 09/07 A2973 RMK AO2 RAE1558B04E16 SLP070 P0001 T00940072 — stamped 16:53Z (10:53 a.m. CST), received 16:56:35Z, report time 17:00Z on Nov 5. A plain hour: no tendency group, no six-hour group. But three SPECIs rode inside it, a second consecutive hour with specials after five silent hours, and between them this hour undid the morning.
The twenty-first rain ended — and it measured. The rain that began at 15:37Z ended at 15:58Z (RAE1558), twenty-one minutes long. At the 15:53Z hourly it had printed P0000, a trace, like the five rains before it. Seven minutes later the 16:00Z SPECI printed P0001: a hundredth of an inch, measured, the trace run broken — the measurement arrived on a special, after the rain had already ended, not on any hourly while it fell. And a twenty-second rain lived and died entirely between specials: B04E16, began 16:04Z, ended 16:16Z, twelve minutes, never in any report's body, known only from the remarks it left on the 16:18Z SPECI and every report after. The hourly total for the hour holding both rains stands at P0001 on all four prints after 16:00Z. At the stamp the body is dry.
The dew point came all the way back down. The trailer had climbed for three stamps — 7.2 → 7.8 → 8.3 → 8.9, +1.7, every step its own. This hour it returned the whole climb in four reports: 8.9 → 8.3 at the 16:00Z SPECI (−0.6, alone, the temperature holding 10.0), a hold at the 16:18Z SPECI, 8.3 → 7.8 at the 16:30Z SPECI (−0.5), and 7.8 → 7.2 at the stamp (−0.6) — back, digit for digit, to the ground it left at 13:53Z. The temperature did not hold its new 10.0 either: it fell alone to 9.4 at 16:18Z, rose alone back to 10.0 at 16:30Z, and fell again to 9.4 at the stamp, an oscillation that ends on its old seven-hourly stand value. At the stamp both halves stepped down together, −0.6 and −0.6 — and the group they landed on is T00940072, the morning's five-printing freeze group, returned digit-for-digit for a second time today, this time by both halves moving at once.
The spread reopened. At the hourlies, 1.1 → 2.2 — doubled, not held. Between stamps it visited 1.7 (16:00Z, the trailer's first step down), 1.1 again (16:18Z, the leader's step down), and 2.2 early (16:30Z, the halves stepping in opposite directions, +0.6 and −0.5 on the same special). The final joint step, equal on both sides, changed the pair and left the spread exactly where the special had put it. The rounded face moved back: 10/09 → 09/07.
IFR lasted exactly one stamp. The 800-ft ceiling did not hold; it lifted at once — 1,400 ft at the 16:00Z SPECI, 2,700 ft at 16:18Z — and the category stepped back to MVFR on the first special and stayed there. Then the ceiling fell back partway: BKN012 from the 16:30Z SPECI onward, 1,200 ft, with BKN020 OVC065 above it at the stamp — three layers again, the highest overcast of the day. The mist is gone from every report this hour; visibility held 10SM throughout, after last hour's swing to 2½SM.
The rise ran an eighth stamp — and steepened again. SLP060 → SLP070, +10, a second consecutive fastest-yet step: the steps now run +4, +3, +4, +7, +7, +5, +8, +10. The altimeter climbed through both specials, A2970 → A2971 → A2972 → A2973.
The gust lived between stamps. The hourly itself prints plain 36014KT — no gust, though the sustained speed, 14 kt, is up from 11. Between stamps the wind was stronger than at either boundary: 34010G18KT at 16:18Z and 34015G22KT at 16:30Z, the strongest sustained speed and gust in today's records on disk. The direction settled north: 330 → 360 → 340 → 340 → 360. PK WND stayed away, an absent twentieth.
Neither signature returned. No AUTO — absent a sixth consecutive hourly. The $ is absent a fifty-third. Neither rode any of the three specials either; eight reports running, across two hours, carry neither.
What did not happen. The rain did not close unmeasured — it measured, a hundredth, on a special after it ended. The dew point did not step a fourth time upward, and it did not hold 8.9 — it fell back the entire climb in a single hour. The spread did not hold 1.1. The ceiling did not hold at 800 ft, and IFR did not persist. The rise did not ease; it steepened to +10. And the hour's strongest wind never printed on an hourly — it exists only in the specials.