Alta Collins precipitation and snow, MOS round 4

Issued Fri 10-30 21:55 MDT · cln_r4_20261030

Issue time (America/Denver; empty for now)

= 2026-10-31 03:55 UTC −6 h −1 h +1 h now

CLN gauge, readings through AGD gauge, readings through

HRRR wrfsubh, runs through HRRR wrfprs, runs through GFS, runs through GEFS, runs through AIFS single, runs through

Cut-offs

Models by hour: amount DA12 × 1, DA11 × 16; wet/dry DW12 × 1, DW11 × 16; SLR DS11 × 17 · Overnight 16:00 to 09:00 MDT: measured 0.22 + expected 0.13 = 0.35 in of water

Hour to (MDT)	Models	p	If wet	Called, p ≥ 0.4	Called, p ≥ 0.3	Expected	Phantom, p < 0.5	CLN gauge	SLR	Expected snow	CLN board	HRRR wrfsubh lead (h)	HRRR wrfsubh runs	CLN lead (h)	AGD lead (min)	GFS lead (h)	GEFS lead (h)	AIFS lead (h)
10-30 22:00	DA12 DW12 DS11	0.43	0.018	0.018	0.018	0.008	0.008	0.00	6.9	0.05	0	1	17	1	20	10	10	10
10-30 23:00	DA11 DW11 DS11	0.45	0.031	0.031	0.031	0.014	0.014	0.00	6.0	0.09	0	2	16	2	80	11	11	11
10-31 00:00	DA11 DW11 DS11	0.43	0.025	0.025	0.025	0.011	0.011	0.01	5.1	0.06	0	3	15	3	140	12	12	12
10-31 01:00	DA11 DW11 DS11	0.50	0.028	0.028	0.028	0.014	0.014	0.00	4.2	0.06	0	4	14	4	200	13	13	13
10-31 02:00	DA11 DW11 DS11	0.38	0.030	0.000	0.030	0.011	0.011	0.00	5.0	0.06	0	5	13	5	260	14	14	14
10-31 03:00	DA11 DW11 DS11	0.44	0.032	0.032	0.032	0.014	0.014	–	5.1	0.07	–	6	12	6	320	15	15	15
10-31 04:00	DA11 DW11 DS11	0.29	0.029	0.000	0.000	0.008	0.008	–	5.6	0.05	–	7	11	7	380	16	16	16
10-31 05:00	DA11 DW11 DS11	0.44	0.037	0.037	0.037	0.017	0.017	–	6.5	0.11	–	8	10	8	440	17	17	17
10-31 06:00	DA11 DW11 DS11	0.44	0.041	0.041	0.041	0.018	0.018	–	6.5	0.12	–	9	9	9	500	18	18	18
10-31 07:00	DA11 DW11 DS11	0.32	0.031	0.000	0.031	0.010	0.010	–	6.4	0.07	–	10	8	10	560	19	19	19
10-31 08:00	DA11 DW11 DS11	0.18	0.022	0.000	0.000	0.004	0.004	–	4.0	0.02	–	11	7	11	620	20	20	20
10-31 09:00	DA11 DW11 DS11	0.18	0.021	0.000	0.000	0.004	0.004	–	5.2	0.02	–	12	6	12	680	21	21	21
10-31 10:00	DA11 DW11 DS11	0.17	0.022	0.000	0.000	0.004	0.004	–	7.0	0.03	–	13	5	13	740	22	22	22
10-31 11:00	DA11 DW11 DS11	0.25	0.022	0.000	0.000	0.006	0.006	–	5.7	0.03	–	14	4	14	800	23	23	23
10-31 12:00	DA11 DW11 DS11	0.38	0.028	0.000	0.028	0.011	0.011	–	5.6	0.06	–	15	3	15	860	24	24	24
10-31 13:00	DA11 DW11 DS11	0.37	0.030	0.000	0.030	0.011	0.011	–	4.4	0.05	–	16	2	16	920	25	25	25
10-31 14:00	DA11 DW11 DS11	0.41	0.028	0.028	0.028	0.011	0.011	–	5.0	0.06	–	17	1	17	980	26	26	26
Total				0.241	0.360	0.175	0.175	0.01		0.98	0

The weights: each source's share in each hour's amount and wet/dry models, the softmax weight ω over the sources in hand (mos/round-4/results_en.md), stacked to 100%. HRRR wrfsubh's share is spread over its runs by lead time. AIFS single has no share: it shifts the chance of a wet hour after the pool. DA12 and DW12 also add AGD's amount already measured in the hour, and HRRR wrfprs feeds only the SLR.

Totals from the start of the issue hour, each 15-minute step included at its end; inches of water, and in the main chart's bottom row inches of snow. Expected (p × the amount if wet) is the forecast total. Its areas split it into likely, from hours with p ≥ 0.5, and phantom, from hours with p < 0.5. Called is each hour as wet, max(amount if wet, 0.01) in, when p is at or above the cut-off, or dry, 0: what an hourly view shows. Its sum is not a total forecast: it counts each called hour as certain and the rest as 0, and in a storm of confident hours it runs above expected. The middle chart has each hour on its own: called at 0.5 against expected, and in replays the gauge. Expected snow is expected × the SLR, the inches of snow on CLN's board per inch of water; like expected, it adds up over hours. In replays the board is CLN's snow board by round 4's rules for it: glitches skipped, a clear counted whole, otherwise each hour's rise from the hour before, and a rise with no water at the gauge in that hour or the hours either side counted as 0. The gauge and the board count only the hours the forecast has. Models: DA for the amount, DW for wet/dry and DS for the SLR, picked for each hour by the sources in hand (mos/round-4/results_en.md); the last hours often take backups, as no HRRR wrfsubh run reaches them within 17 hours yet. HRRR wrfsubh runs: up to 10 runs in hand, up to the one picked, lighter the earlier, each the mean of its 4 points. Each MOS hour is split into 15 minutes as the HRRR wrfsubh run with the shortest lead time to it splits it, evenly without one. Each source lists its runs or readings in hand at the issue time that still reach a forecast hour, runs by their UTC start and readings by their local time; the forecast uses the source up to the one picked, and changing the issue time sets every source back to its default. In a replay, a source's default is its last run or reading at least this long before the issue time: the CLN gauge and AGD 15 minutes, HRRR wrfsubh and HRRR wrfprs 50 minutes (the whole run), GFS 5 hours, GEFS 6 hours and AIFS single 5 hours. Built in 1.02 s.