Curious Iowa: How are daily normal temperatures calculated, and how reliable are they?

The normal high temperature on Dec. 22 in Cedar Rapids, according to the National Weather Service, is 31 degrees. The normal low is 15 degrees, and the average precipitation amount is 0.73 inches.

The actual weather experienced today might vary from those averages — likely with a high close to 40 degrees, a low close to 32, and no precipitation as of KCRG-TV9’s 10-day weather forecast Friday — but the numbers from today’s recorded weather won’t be added to the data used to calculate those normal highs and lows until 2031.

So, how are those numbers calculated, and how far back does the data that’s used to find temperature and precipitation averages go? Those are the questions that one reader sent to Curious Iowa — a series from The Gazette that answers readers’ questions about the state and how it works.

How are weather averages calculated?

According to KCRG-TV9 chief meteorologist Joe Winters, the averages used in KCRG-TV9 weather reports — which The Gazette relies on for its daily weather information — are based on a 30-year period ending with the last completed decade. Currently, that means precipitation averages and normal highs and lows are based on numbers from 1991 through 2020. In 2031, the data will be updated to include numbers from 2001 through 2030.

This is common practice among meteorologists — rather than calculating the averages and normals from all the years of reported data available in an area — both because it provides consistency between different locations when comparing climates, and because using a 30-year period of time can account for more recent changes in the climate to give a more accurate idea of what weather can be expected in the future, Winters said.

“I don’t know of any calculation that takes the entire period of weather record,” Winters said.

The 30-year period also is used by climatologists, like Iowa State Climatologist Justin Glisan, to measure change in temperature over time and to study how the climate changes on a large scale.

“We’ve seen more rapid atmospheric warming later in the 20th century and in the 21st century than we did in the 1800s. That’s why the averaging period is so important,” Glisan said. “Fifteen to 30 years is the sweet spot when we're talking about averages. If you take a 100-year average for the 20th century, it's going to filter out the warming that we've seen at the tail end of the 20th century, so it kind of smooths out the temperature increase overall.”

How much weather data does Iowa have?

Although normal temperature highs and lows are calculated from the last 30 years of data, Iowa’s weather data goes back much further than that.

Daily highs and lows and precipitation amounts have been recorded in different parts of the state for more than 100 years. In the Cedar Rapids area, weather observations have been recorded since 1953 at The Eastern Iowa Airport and since 1893 at a weather observation station in Marion. The data is recorded each day at 7 a.m. by a dedicated observer, who records information from rain gauges and thermometers that show the highest and lowest points reached in the previous 24 hours.

Even before the days of digital measuring instruments, these numbers were marked automatically and recorded daily from thermometers designed to leave a mark at the highest and lowest levels reached during a day. Precipitation still is measured without a digital system, instead relying on a series of buckets that tip over when filled. This gives observers a clear way to see how much water is captured in a day, according to Glisan.

“The instrumentation has changed that take the observations, but the quality control algorithms and the quality control that we do on the data has not changed. It’s only improved,” Glisan said. “There’s a quality control algorithm that goes into the weather observations. So, if we notice a station is running a warm bias, or is running warmer than surrounding stations, that station is checked to make sure that the instruments are working properly.”

Record high numbers and record low numbers for each day are reported based on all the data that an area has on record, rather than just the 30 years used to calculate averages.

How reliable are daily normals and averages as a predictor for actual weather?

Although using 30 years of data to calculate the daily normal temperatures means the numbers better reflect recent weather trends, that doesn’t mean you can always count on those numbers to be accurate in any given year.

David Flory, an Iowa State University professor who specializes in meteorology, said that while average temperatures can give a general idea of what temperatures might be several months from now, weather systems can change significantly from year to year. Those changes can be related to larger weather patterns that can be predicted further out — like a La Nina or El Nino annual climate pattern that will result in colder or warmer weather — but can also come down to daily changes.

“It’s more variable from day to day. The saying is, climate is what you expect, weather is what you get,” Flory said. “Can you really use it as a good predictor for a wedding a year down the road? You can get some idea of what you might expect, but don’t think that that’s the ground truth.”

Have a question for Curious Iowa?

Tell us what to investigate next.

Comments: (319) 398-8328; emily.andersen@thegazette.com