The hundred-degree day count, and what it measures

A hundred-degree day is the simplest statistic Phoenix keeps and the easiest to misread. It is not a temperature and it is not a forecast. It is a count of calendar days on which the high reached at least 100 °F, and the number it produces for a normal year carries real information about the shape of a Phoenix summer, as long as you know which question it answers.

This page is about that number: where it comes from, how it accumulates through the year, and the three things it does not measure.

What a hundred-degree day is

The definition is a threshold, not an average. A day counts if the maximum temperature at the observing station reaches 100 °F or higher. That is the whole rule. It does not matter whether the high was 100 °F or 118 °F; both are one day.

That is worth stating plainly because the threshold is a human convention rather than a physical boundary. Ninety degrees would make the season longer. A hundred and five would make it shorter and more severe. The United States keeps 90 °F, 100 °F and 105 °F counts for exactly this reason: each answers a slightly different question, and none of them is the temperature itself.

Where the 111.3 comes from

The figure for Phoenix is 111.3 days, and it is a normal rather than an observation. It comes from the NOAA NCEI U.S. Climate Normals for 1991–2020, measured at Phoenix Sky Harbor International Airport, station USW00023183. A normal is a thirty-year average, recomputed once a decade, so 111.3 is what a typical year in that period produced, not what next year will produce.

The monthly normals add up to the same total. Each month carries its own normal count of days above 100 °F, and the twelve of them sum to 111.3. The station record behind those normals is long (daily observations at Sky Harbor run back to June 1933) which is what makes a thirty-year average meaningful rather than a guess.

None of that makes 111.3 a number to plan a specific date around. It is a description of a climate, and the month-by-month guide to how hot Phoenix gets is the companion to it.

How the year fills in

The count is not spread evenly. It barely exists for the first third of the year and then arrives almost all at once, which is what the cumulative curve below shows.

Put to numbers, the shape is stark. Nothing accumulates through January, February and March. April contributes one day. May contributes nine, for a running total of ten by the end of the month. June adds 24.6, taking the year to 34.6. July adds 28.5, and August 27.1: together more than half the year’s total in two months. September adds 18.2, October adds the last 2.9, and the number is finished.

Two counts that look alike

There are two figures close to 111 in the Phoenix record, and they are not the same statistic.

The first is 111.3 days, the annual normal: the average number of days in a year on which the high reaches 100 °F.

The second is 111 days, the length of the window from 31 May to 18 September in which the normal daily high is at or above 100 °F. That is a run of dates, read off the daily normals day by day; the first figure is a yearly total.

They land within a day of each other, which is a coincidence of this climate rather than a rule, and the two can drift apart in any year that is not normal. When a page says Phoenix has “about 111 hundred-degree days”, it is usually quoting the first figure and sometimes meaning the second.

The number is an average, so it hides the years

A normal is the middle of thirty years, and the middle is not where any single summer sits. The county’s report on 2024 records 113 consecutive days at or above 100 °F, from 27 May to 16 September, in a summer it describes as the hottest on record. Consecutive is a different measure from a count, but the two point the same way: a year can run well past the normal without the normal changing.

That is the reason the Outdoor Event Heat Check works from the station record rather than the monthly averages. A monthly normal tells you what a typical July is. It does not tell you what your date has done, and an event gets one attempt at its date.

What the count does not measure

Three limits are worth keeping in view, because the count is quoted so often that they get lost.

It counts days, not intensity. The normal hottest day of the Phoenix year is 4 July, at a normal high of 106.9 °F. A 100 °F day and a 107 °F day both add one to the total, and anyone who has spent both in the sun knows they are not the same day.

It says nothing about the nights. The count is built from daily maximum temperatures and ignores the minimum entirely. On the same 1991–2020 normals, Phoenix has 84 nights whose low stays at 80 °F or above, from 20 June to 11 September, and 141 nights at or above 70 °F. Those nights are the reason a summer can be dangerous at 3 a.m., and the day count is blind to them. The page on why Phoenix nights stay hot takes them on directly.

It is one station’s number. The figure is measured at a single thermometer, at Sky Harbor, which is an urban airport site. A count from a different part of the Valley would not be the same count, and 111.3 describes that station’s climate rather than the whole metropolitan area’s.

Why it is still worth keeping

For all that, the count answers a question nothing else answers as cleanly: how long is the hot season, and where inside it does the weight fall. Knowing that half of the year’s hundred-degree days arrive in a two-month core is useful for anything that has to be scheduled (a move, a roof, a training plan, a wedding) and knowing that the count starts in April and ends in October is useful for anything that has to be stored or shipped.

Set against the Heat-Safe Hours Planner, it becomes a two-part answer: how many days of the year are hot, and which hours of those days a person can safely work in. The count gives the season. The planner gives the hour.