The west coast of the United States got a bit of a heat spell in June of 2021. In fact, it broke a number of records. Some weather stations on the west side of the Cascade Mountains recorded temperature highs of 116°F, or 47°C, where the local climate is usually mild. Heat spells like this always raise the question: are hot days like this to be expected every now and then? The frequency distribution chart above illustrates temperature highs recorded at a historical weather station, from 1891 through 2020, and these years represent the duration of record keeping at that weather station. The highest temperature recorded for that period is 106°F, or 41°C. The data available for this weather station shows 41 days above 100°F, 3 days above 105°F, and 0 days above 110°F. Use your mouse cursor to explore the temperatures by calendar day in the following interactive charts.

Frequency distribution for years 1891-2020 in °F

Frequency distribution for years 1891-2020 in °C

You have probably noticed the horizontal bands. It is more noticeable in the Fahrenheit chart because one degree Fahrenheit is 1.8 times more precise than one degree Celsius, and both charts have their frequency plotted out in one degree steps.

The pattern of horizontal banding occurs every 5 degrees Celsius, Fours and nines seem to be frequently rounded up to fives and tens, as though the station keeper were employing a rounding bias when reading the thermometer. Besides that, values in the database are rounded to a specific pattern of decimal values: 0.0, 0.6, 1.1, 1.7, 2.2, 2.8, 3.3, 3.9, 4.4, 5.0.

The measurements were originally taken in Fahrenheit as whole numbers, and then converted to Celsius, in order to be stored in the Global Historical Climate Network (GHCN) database. This is apparent because the database values, converted to Fahrenheit, are very close to integer values. In fact, the database values, converted back to Fahrenheit, are all within 0.09 of being whole numbers. Fahrenheit values, rounded to the nearest integer, then converted to Celsius, then rounded to the nearest tenth, will correctly match the database values.

This frequency distribution chart represents one of the different ways that weather data can be represented using 3D charts. Alternatively, Climate Binge can also produce a yearly chart, having x as the calendar day, y as the year, and z as the temperature. Climate Binge is built in the spirit of citizen science, where you do the science, and you draw your own conclusions! Head on over to the charting tool to choose your own weather station and create your own 3D weather chart.