Popular science is full of hidden quantities. The boiling point of water, 100°C/212°F, assumes one standard atmosphere of pressure (1.013 bar). The boiling point varies proportionally with air pressure. At 10000 ft (3048 m) elevation, for example, water boils at 89.5°C/193°F (which is why performance climbers need pressure cookers to boil food). When you read or hear about the decibel level of sound, such as at an airfield or in a domed stadium, the decibel count is power (exponents) ratio of sound pressure level against a reference (usually 20 micropascals, uPa, considered the lower threshold of normal human hearing). That reference pressure is a type of datum: a fixed reference for other measurements.
Datums permeate science. In mapping, establishing datums such as the shape of the earth (the “reference ellipsoid”) and “sea level” is critical for giving useful depths and elevations. (Establishing “sea level”, otherwise known as the “geoid”, is a complicated and very important process.) The same is true in climate science. There are reference quantities used as the basis for comparison against other measured values, and these reference quantities are called “normals”.
Frequently, when looking at a map of temperature or precipitation anomalies, the map will state the datum for comparison (“pre-industrial”, i.e. 1750-1800; 1900-2000; etc.). But the climate is dynamic enough—variables such as temperature move steadily away from previous baselines—that resetting the baseline, resetting the datum, is useful. So the World Meteorological Organization (the WMO) has set up the rule that countries must publish new climate normal—averages of the previous 30 years’ meteorological data—every ten years.
This is to provide a rolling window of statistically useful comparison. 30 years is enough time to account for climate trends (the meteorological definition of climate is the 25-year average of weather), but short enough that the average values are closely comparable to the present day. The US’ latest set of climate normals, spanning 1990-2020, was published in May 2021. These normals will serve as the basis for meteorological anomaly maps for the next ten years.
The normals are calculated carefully from weather stations across the country, accounting statistically for missing data points (if an instrument went down for any length of time) and biases (such as urban heat islands or proximity to other stations). Plots of the different normals over time, compared against a longer datum (such as the 20th century) are very instructive: long-term weather averages compared to a much longer-term average: climate vs climate: gives a very robust view of climate change.
Tomorrow: resetting the year-by-year climate annals, beginning with 1998.
Be brave, and be well.