Jerry Killingsworth, meteorologist with the National Weather Service office in Goodland, says he doesn’t trust data being recorded by a weather station on private land and can’t check on the equipment because of federal rules regarding COVID-19. (Submitted to Kansas Reflector)
Temperatures recorded at a western Kansas weather station in late September captured the attention of a researcher looking for the first measurable snowfall in the Midwest.
Analysis by Brian Brettschneider, a research physical scientist with the National Weather Service Alaska region, led him to a cooperative observation station 13 miles north-northeast of Tribune in Greeley County. That station had recorded snow accumulation of 0.3 inches on Sept. 9, then measured daytime high temperatures of 102 degrees on Sept. 25 and 26.
Those measurements have raised concerns about consequences of the COVID-19 pandemic on data accuracy, frustrating climate researchers. Across the country, dozens of manually operated cooperative sites are missing data or providing unreliable data because federal employees following health guidelines haven’t had access to the equipment.
Brettschneider entered his specific search parameters in the Global Historical Climate Network, as well as the Applied Climate Information System. He then was disappointed to learn the Tribune weather station is unreliable.
“I had access to the complete GHCN daily data for the U.S. and Canada, which consists of 18 million daily records,” Brettschneider said. “I have some programs written to cull through certain things; this was an unusual query.”
Jerry Killingsworth, meteorologist and manager of the cooperative weather station program at the NWS office in Goodland, said the Tribune 13 NNE station has been reading inconsistently for several months.
“My colleagues and I agree that it’s not a trustworthy site for high temperature readings,” Killingsworth said.
The station is one of a handful of cooperative observation stations in his region that are sending unreliable or incomplete data. Killingsworth said in most cases, these stations are on private property, and the data collection units are usually inside a person’s home. Currently, federal employees with climate and weather agencies are barred from visiting these stations because of health concerns over COVID-19.
“Climate records are being impacted, and the issue is far-reaching,” Killingsworth said.
Amy Fritz, manager of the NWS Cooperative Observer Program — which provides observations for the official climate record from more than 8,000 official sites — said 98 out of the 6,375 manually operated cooperative locations were missing data as of the end of July. It has not been determined how many of those sites are in Kansas.
Killingsworth said he would not have caught the data inconsistency if it had not been brought to his attention. He has been marking a lot of the high temperature data from that weather station as “missing,” as it is unreliable.
“I don’t want bad data going out, and I can’t go out to the site to check on it or fix it, so my only option for now is to set the data as ‘missing,’ ” Killingsworth said.
Data from the cooperative observation stations goes through multiple avenues of quality control, starting at a local level with NWS offices. Monthly data reports are reviewed by a meteorologist in office, then are sent off to the National Center for Environmental Information in North Carolina, where they undergo more quality control and review.
Fritz said missing data because of equipment issues will be resolved as soon as researchers are able to safely access these locations.
Killingsworth said he does not know when he will be able to visit the Tribune 13 NNE station. He said it is likely a simple fix, such as a sensor issue or the equipment being placed too close to a building.
“The coop program is hurting,” Killingsworth said. “This is important data, and it’s frustrating that we can’t trust it, and can’t fix the problem.”
Killingsworth said he has noticed a weather observation station in Cheyenne County, Colorado, having the same issue with extreme daily high temperature readings. The latest point in the year for a triple-digit daytime high recorded by the Goodland NWS office was on Sept. 14, 1895, with a high temperature of 101 degrees.
Brettschneider said he was conducting his segment of analysis on a freelance basis, not affiliated with his official NWS work, and that he enjoys pulling and sifting through weather data to find patterns or interesting points. He said he feels strongly about communicating climate science and data in a digestible way, but to do so accuracy is of utmost importance.
“If you don’t get good information out there that people want, that space could be filled by bad information,” Brettschneider said. “You have a duty to occupy that space with good information.”
Our stories may be republished online or in print under Creative Commons license CC BY-NC-ND 4.0. We ask that you edit only for style or to shorten, provide proper attribution and link to our web site. Please see our republishing guidelines for use of photos and graphics.