Were More Than 3,000 Heat Records Actually Set in June?

BY NewsHour Labs  July 6, 2012 at 5:11 PM EDT

Here at the NewsHour, the record heat has hit close to home. Our managing editor Tom Kennedy, for example, spent six of the last nine days of 95-degree-plus heat without power. The thermostat peaked at 90 degrees inside his Maryland home, forcing his family and the 12-year-old golden retriever to sleep in the basement so they wouldn’t overheat. And this is a common story.

But just how extreme is the heat? A widely reported headline this week announced that 3,215 high temperature records were matched or broken across the U.S. in June. But the PBS NewsHour’s record temperature heat widget only reported 743 broken heat records–quite a discrepancy.

Welcome to the murky challenge of reporting on temperature extremes, where some heat records are more notable than others, some monitoring stations have been around longer than others, and every skeptic is an armchair climate scientist.

We became aware of all of these issues over the course of the construction of our widget, so we reached out to the National Climatic Data Center–the scientists who report the records–for guidance. Here’s what we learned:

Not all Broken Records Are Created Equal

The National Oceanic and Atmospheric Administration collects temperature data from a network of 931 automated “official” weather monitoring stations, and from a broader network of more than 10,000 volunteer weather watchers who meet certain standards. These volunteer weather stations provide valuable data to scientists who track temperatures across the United States. However, grouping record highs recorded by these stations into one lump runs the risk of oversimplifying the science, so our widget only reports records measured at the official stations.

For instance, two weather stations that are close in proximity might experience record high temperatures on the same summer day–let’s say 108 degrees at both locations. But suppose one weather station’s temperature archive dates back to 1905, while the other only started capturing temperatures in 1975. The newer station might have missed a record-hot summer in 1954 when temperatures reached 111 degrees. Is it accurate to report that two record temperatures were set, or only one? When reporting record temperatures, NOAA only includes stations with at least 30 years of data, but most of the official stations have much longer histories. NOAA says that using official stations over volunteer stations ensures that the comparison is between similarly sized sets of stations year-over-year.

Weather Station Overlap

PBS NewsHour Managing Editor Tom Kennedy sets up a solar panel which collects energy to use at his home. He and his family haven’t had power for six days and have been enduring the record heat in Washington, D.C.

NOAA’s temperature-gathering network is so large that measurements are sometimes taken from virtually the same location. For example there are two volunteer weather stations less than a mile apart in Contra Costa County in the Bay Area. Official stations are much less likely to be located in such close proximity.

Delays in Data Reporting (Avoiding Snail Mail)

We also chose to avoid counting the volunteer stations because the process of collecting data from them is more of a manual process. The head of NCDC’s Climate Monitoring Branch said in an email to us last year that measurements from these stations often arrive several days or weeks after they were taken — sometimes by U.S. mail.

Continue to follow our widget tracker as the summer continues. The chart below shows that 2012 is currently on track to be one of the most record-breaking years documented. So, we doubt this widget will become irrelevant anytime soon.

Also, let us know if you have questions or suggestions in the comments field below.

Travis Daub, Saskia De Melker, Vanessa Dennis, Jenny Marder and Justin Myers contributed to this report.

Correction: An earlier version of this story incorrectly reported the neighborhood where our managing editor lives.