Faribault 7-day forecast

Each piece of a card, shown with tomorrow's own numbers.

The bar

• Normal
• Likely low
• Likely high
• Our forecast
• Co-op record low and high

Where the low and high should land about 8 times in 10 (likely), against the normal low to high for the date (NOAA 1991-2020) and the co-op stationCo-op station A National Weather Service station kept by a volunteer, who reads temperature and precipitation at home or work. National Weather Service glossary's record low and high for it (its daily series runs 1895 to 2011). A line under the bar when the forecast is within 2° of that record or beyond it. The normal and record lows are for the whole calendar day, not the night after it, so they don't measure quite the same thing as our low.

Models

Models on the low and rain: Sure. Sure on the low: GFS, GEM, NBM and ICON within 1°. Sure on rain: all 6 models at 0.05 in. or more.

How close the models are on the day's low and on rain. Beside each word: the models it rates and how far apart they are. The low: how close GFS, GEM and NBM are (and ICON through day 5). Sure within 3° for days 1 to 3 and 6° for days 4 to 7, Fairly sure within 8° and 11°, Not sure past that. Rain: Sure when every model agrees on rain or no rain (0.05 in. or more), Fairly sure when one model disagrees, Not sure on a split. Rain's models: HRRR, NAM, GFS, Euro (9 km), ICON, GEM and NBM. A day's word is the lower of the two. Model spread doesn't predict how far off the high will be, so the high gets its range, not a word.

Rain

Rain most likely 0.20 to 0.65 in., mostly morning to evening.

• Morning81%
• Afternoon83%
• Evening89%
• Overnight76%

The likely amount: the middle half of the ensemble members' totals among those with rain, to the nearest 0.05 in., shown with a chance of 20% or more; and when: "mostly morning to evening" names the parts of the day whose NWS hourly chance is within 10 points of the highest, when the highest is 30% or more, all four are known and not all are that close. The chances are in the table below it, ours and the NWS's; on a card without that table (Today, and day 7 before the NWS reaches it) the line gives the NWS chance for the day and the night after it (ours, the ensembles', when the NWS has none). Ours is the share of GFS and Euro ensemble members with 0.01 in. or more that day, each ensemble counted equally. The bars: the NWS hourly chance, highest in each part of the day (morning, afternoon, evening) and the night after it (overnight, midnight to 6 a.m.). When every part is under 10% and our own chance for the day is too, one line stands in for the bars, naming only the parts the hourly forecast covers to their end: "Dry: under 10% chance all day (NWS)." when it covers all four, "Dry morning to evening: under 10% chance (NWS)." on a day it reaches only in part. On the Today card, which has no chance of ours, when the NWS chances for its periods are under 10% too, "Dry the rest of today: under 10% chance (NWS)." stands in for both.

Ours and the NWS

Our forecast and the National Weather Service's

For	Ours	NWS	Differs
High	55°	51°	NWS 4° cooler
Low	47°	41°	NWS 6° cooler
Rain	100%	89%

Our high and low beside the National Weather Service's high for the day and low for the night after it, and both chances of rain. A row is marked when the high or low is 4° or more apart, or on rain when one is at 50% or more while the other is under 30% (Split). When the two chances are 30 points apart or more, a line under it says why they can differ: the NWS's is its forecast for Faribault's spot over its 12-hour day and night periods; ours counts ensemble runs over the calendar day.

How it moved

Our forecast for the day as it stood at each daily save (9:40 a.m. CST), from the models' runs on hand then, the oldest on the left, with today's model weights, ending with now: the card's own forecast, from the newest runs; a sentence when it moved 4° or more.