Outdoor Event Heat Check
Choose a date and a temperature and see how often Phoenix has reached it on that day, from the station’s daily records, so an outdoor event can be planned against the odds rather than the average.
The date you are considering, not today. Setting it to a date you have already chosen is the point of the tool.
A day counts when its maximum reaches this figure at any point. An evening event inherits the day's maximum whether or not it is still that hot when guests arrive.
Phoenix Sky Harbor, 93 calendar years, 32,201 daily observations, 1933-06-01 to 2026-10-07.
Never recorded
0%
of years reached 100 °F on this date
In 87 years of observations, Nov 1 has never reached 100 °F at Phoenix Sky Harbor. That is not a promise about this year (the record is 93 summers long, and a hotter one is always possible) but it is the honest answer from the data.
- Years reaching it
- 0 of 87
- Normal high on this date
- 81.3 °F
- Hottest ever recorded
- 96 °F
| Temperature | Years reached | Share of years |
|---|---|---|
| 90 °F | 4 of 87 | 4.6% |
| 95 °F | 1 of 87 | 1.1% |
| 100 °F | 0 of 87 | 0% |
| 105 °F | 0 of 87 | 0% |
| 110 °F | 0 of 87 | 0% |
| 115 °F | 0 of 87 | 0% |
| 120 °F | 0 of 87 | 0% |
This tool asks a different question from every other heat tool on the site. The Heat-Safe Hours Planner works from climate normals and answers "what is a normal day like". This one ignores what is normal entirely and counts what happened, because the person using it is usually choosing a date for something they only get one attempt at.
Where the numbers come from
Every figure is read from the National Centers for Environmental Information's Global Historical Climatology Network (Daily) record for station USW00023183, Phoenix Sky Harbor International Airport. That record begins on 1933-06-01 and runs to the present, and this build counted 32,201 daily maximum temperature observations across 93 calendar years.
Each observation is filed against its calendar date (not its day of the year, its date), so 3 July 1974 and 3 July 2019 land in the same bucket and 4 July lands somewhere else. Each bucket then records how many years it holds and, of those, how many reached each of seven thresholds: 90, 95, 100, 105, 110, 115 and 120 °F. What the page shows is that count divided by the years.
Values that NOAA has flagged as failing its own quality checks are left out. That matters in one direction only: it can make a percentage slightly lower than a count-everything approach would, and it can never make this site report a hot day the agency had already discarded as suspect.
Why the date index looks odd
The data is stored as one series per threshold, indexed by day of year. Because 29 February has to live somewhere, the index is built through a 29-day February: 1 January is index 0, 29 February is index 59, and 1 March is index 60 in every series, leap year or not. A common year simply leaves index 59 unused. The alternative (an ordinary day-of-year count) would shift every date after February by one day in leap years, which is exactly the class of bug that produces a confidently wrong answer.
How to read a percentage
A figure of 97% means the date reached the threshold in 97 years out of every 100 on record. It does not mean a 3% chance of a cool day: the years it misses are usually misses by a degree or two, not by ten. The tool prints the normal high and the all-time record high for the same date next to it, and those two together describe the spread far better than the percentage alone.
The one-in-N phrasing is deliberately absent from the readout. "One year in twelve" sounds like a lottery and reads as reassurance, when what it means is that in one year out of the last twelve your event was a hundred-and-ten-degree event. The percentage and the raw count are both shown, and the reader can decide which framing is honest for their decision.
What the bar column shows
The table lists all seven thresholds for the chosen date in one place, with the count and the share for each. Reading down the column is the useful part: the drop between one row and the next is the price of moving your threshold up five degrees, and it is rarely a gentle slope. For a June date the share at 100 °F might be 99% and the share at 110 °F 12%, which is a much more useful thing to know than either number on its own.
- NOAA National Centers for Environmental Information: Global Historical Climatology Network, Daily (GHCN-Daily), station USW00023183, Phoenix Sky Harbor
Every count on this page. The station's daily maximum temperature record from 1933-06-01 to 2026-10-07 (32,201 daily observations across 93 calendar years) read once, counted by calendar date, and shipped to the page as counts.
- NOAA National Centers for Environmental Information: U.S. Climate Normals 1991–2020, daily summaries, Phoenix Sky Harbor (USW00023183)
The normal high printed beside the record high, so the two can be compared. The odds themselves do not use the normals.
Every figure this tool uses was read from the source named above on 10 October 2026. Anything still waiting on a check is listed in the site's data notes rather than shown here as if it were settled.
Before you rely on this
This tool works from climate normals and published guidance, not from live weather. It describes a typical day for the month or date you chose, which is useful for planning and useless as a forecast.
- The record begins on 1933-06-01 at this station. Anything before that is not in these counts, and a date with fewer years behind it (29 February most obviously) carries a wider error bar than the percentage suggests.
- Phoenix has warmed across the record. A percentage that averages 1933 to the present understates a recent year and overstates a 1950s one, because the later decades contribute more hot days to the same denominator. The counts are honest about the whole record; they are not a trend line.
- This is one station. Sky Harbor runs warmer than outlying parts of the Valley at night and slightly cooler than the deep southwest in the afternoon, so a date that is marginal here can go either way ten miles out.
- A daily maximum is not what your event experiences. A 4 p.m. ceremony and a 7 p.m. one on the same date share this number and do not share the weather.
- Thresholds are counted at 5 °F steps and the tool snaps your input to the nearest one. Interpolating between them would produce a figure no observation supports.
- Values NOAA has flagged as failing its quality checks are excluded, so a reading the agency discarded as suspect is not counted even where it would raise the total. That makes these percentages slightly conservative rather than slightly wrong.