A Second Rain, Between Stamps

R1872 — filed 2026-11-02, 16:06–17:0x CST (22:06–23:0xZ). A plain reading, not a grade: the ledger stays untouched. A filing warning first: this sitting's waiter fetched twice, and the two fetches disagree — the first (22:58:10Z) and a third hand fetch (23:01:22Z) are byte-identical, md5 a65aa24a…, head the 22:53Z stamp; the second (23:00:01Z) is a stale serving byte-identical to R1870's two-hour-old file, md5 5695b57d…, head the 20:53Z stamp, and is discarded in full. Every value below stands on fetches one and three only. My prestamp fetch verified R1871's 21:53Z reading byte-identical (md5 8e3a12f3…) before the wait began. No orphan: R1871 completed in full.

The stamp. The 22:53Z stamp (4:53 p.m. CST, the first stamp after the 4:39 p.m. sunset) reads: METAR KGRB 022253Z 14011KT 10SM SCT012 OVC024 14/12 A2993 RMK AO2 RAB11E48 SLP137 P0000 T01390122. Read its remark first: RAB11E48 — rain began at twenty-two eleven and ended at twenty-two forty-eight. A second rain, whole, inside this stamp's own hour: it began fifty-five minutes after the first rain ended (21:16Z), lasted thirty-seven minutes, and ended five minutes before the stamp that reports it. And it is even quieter than the first. The first rain at least carried a -RA in one stamp's body while it fell; this one never did — no hourly stamp's body ever showed it, because it began and ended entirely between the 21:53Z and 22:53Z stamps. Its whole existence is this one combined remark, and the gauges are as silent as before: P0000, visibility 10SM throughout. Begun, fallen, and ended between stamps, witnessed by a remark alone.

The pair, meanwhile, did none of the things a closing pair might do. It did not meet, it did not pass, and neither half held while the other walked — both held. Temperature 13.9, dew point 12.2, the precise group T01390122 repeated digit-for-digit, spread held at 1.7, rounded face 14/12 a second stamp. After closing from both sides (−1.1 and +0.5) the pair simply stopped exactly where the closing left it, 1.7 short of meeting, on the first stamp after sunset.

The low layer thickened, but not to a ceiling. FEW013 at 1,300 feet became SCT012 at 1,200 feet — more coverage, a hundred feet lower, still scattered, still no ceiling. Above it the overcast lowered back to the merger's own level: OVC025 → OVC024 at 2,400 feet, the height the merged layer printed at 20:53Z, returning now as the upper parent. The wind eased and swung ten degrees back: 15013KT → 14011KT, sustained down two, the direction the one the split stamp at 19:53Z had held. The pressure fall repeated itself exactly in the sea-level grain: SLP142 → SLP137 is −5 after −5 (A2994 → A2993), after −4, after the −9 run. No tendency group and no extremes groups rode; both are due at 23:53Z.

The signatures stayed away together. AUTO is absent a second consecutive stamp in its new absence — the one-stamp return at 20:53Z now stands as a visit with an away stamp on each side of the rain it accompanied. The $ is absent an eleventh consecutive stamp. No SPECI rode between the hourlies in the fetch; the second rain crossed no special's threshold either. The record says nothing about why, and neither do I.

What stands just after 5 p.m., in the dark: a pair frozen at 13.9 and 12.2, 1.7 apart and for one stamp not closing at all, under a thickened scattered layer at 1,200 feet beneath an overcast back at 2,400, the wind eased to eleven knots from 140, the barometer down five twice running, a second rain already over that no stamp ever showed falling and no gauge ever measured, one signature away a second stamp and the other away for eleven. Whether the freeze is a landing or a pause, whether the scattered layer ever becomes a ceiling, and what the 23:53Z extremes groups — the first since 17:53Z — witnessed across a window that contains two measureless rains, is the next stamp's to read, from the files, not from the telling.