The FCC recently auctioned spectrum in the 24GHz band under controversial circumstances, as experts from other federal agencies warned that cellular transmissions in that band may significantly reduce the accuracy of weather forecasts.

When asked about the controversy at yesterday’s Senate Commerce Committee hearing, Pai said that data provided by NASA and the National Oceanic and Atmospheric Administration (NOAA) is faulty. He also criticized the agencies for raising concerns “at the 11th hour.”

Calling NOAA’s study “fundamentally flawed,” Pai said, “For example, it ignores the fact that 5G will involve beamforming, essentially adaptive antenna arrays that will more precisely send 5G signals—sort of a rifle shot, if you will, instead of a shotgun blast of 5G spectrum.”

Water vapor and weather sensors

The 24GHz-band spectrum purchased by AT&T, T-Mobile, and other carriers ranges from 24.25GHz to 25.25GHz.

Water vapor emits radiation at 23.8GHz. NASA and NOAA scientists are worried that sensors won’t be able to pick up these faint signals because of interference from 5G transmissions in the adjacent 24GHz band; a US Navy memo also raises this concern. The agencies say this could reduce the accuracy of forecasts for hurricanes and other storms.

Pai stressed that the 5G spectrum “is separated from the passive weather sensors in question by over 250MHz.” The FCC’s proposed emission limits for the spectrum are “appropriate for protection of passive weather sensors,” he said.

Pai also told Congress yesterday that there are already “40,000 fixed and microwave links” operating in the band just below 23.8GHz. Despite that, he said, “there’s never been a reported case of interference” with weather sensors.

“Over the last two and a half years we’ve patiently waited for a validated study to suggest that our proposed [emission] limit is inappropriate. We’ve never gotten such a validated study,” Pai said.

Requiring lower-power emissions in 24GHz spectrum could make the airwaves “unusable for 5G domestically,” Pai said.

NOAA projects 77% data loss

FCC rules allow out-of-band emissions of -20dB, a level that NASA and NOAA say is too high. This “would result in roughly a 77 percent data loss from our passive microwave samples,” bringing accuracy down to levels not seen since about 1980, NOAA Assistant Secretary Neil Jacobs told members of Congress in a House hearing on May 16.

For hurricane forecasts, this would reduce forecast lead times “by roughly two to three days,” Jacobs said.

“It’s incredibly important—it’s a critical dataset for us,” Jacobs said. NOAA’s data indicates that an out-of-band emission level of about -50dB instead of -20dB would solve the problem, “result[ing] in roughly zero data loss,” Jacobs told lawmakers.

At the FCC-approved levels, by contrast, data loss could be “large enough to prevent us from meeting our mission requirements with the future JPSS system,” Jacobs said, referring to NOAA and NASA’s planned Joint Polar Satellite System.

Jacobs told Congress that the “Department of Commerce supports 5G,” and that he is “optimistic that we can come up with a solution where passive microwave sensing and 5G can coexist.” (NOAA is part of the Department of Commerce.)

Jacobs didn’t say when the NOAA study will be released publicly but said the agency has been “going back and forth” with the FCC and NASA about the details. Pai is not convinced. He told lawmakers yesterday that the FCC asked for input from federal agencies on the 24GHz band in 2017 and didn’t receive any credible concerns.

“One of our federal partners said they had a study [showing potential interference] in 2017. We then worked with them as well as other federal stakeholders to identify some serious flaws in that study. That study was ultimately withdrawn,” Pai said.

When the 24GHz auction began on March 14, 2019, “we had not had any validated study that showed that our protection limit was inappropriate,” Pai said. He continued:

At the 11th hour, however, one of our federal partners suggested there was such a study. We asked to work with them, get some insight into what their study was showing them. Until May 10, our staff never got the code that was necessary for evaluating that study. So, far from it being peer-reviewed, it wasn’t even reviewed by some of the sister agencies.

Pai said that “the assumptions that clearly underlay that study were so flawed as to make the study, in our view at least, meaningless.”

Pai also said that the Department of Commerce’s position could undermine the United States’ ability to negotiate at this fall’s World Radiocommunication Conference, “where the overall world limit for the 24GHz band among other 5G bands will be set.”

“Unfortunately, one department has been very active in trying to undermine the United States’ position in these international negotiations and make it more difficult for us to free up spectrum in 5G,” Pai said.

While Pai accused NOAA of using bad data, Pai himself has used bad data to bolster his claims that his policies have increased broadband deployment.

“The situation is embarrassing”

When contacted by Ars about Pai’s criticism today, the Department of Commerce referred us to Jacobs’ testimony from last month but provided no further comment.

Pai’s concerns were met sympathetically by Republicans on the Senate Commerce Committee. The NASA and NOAA objections “make no sense whatsoever,” Sen. Ron Johnson (R-Wis.) said.

Senate Democrats have criticized Pai for moving forward with the auction, which was completed on May 28. FCC Commissioner Jessica Rosenworcel, a Democrat, said at yesterday’s hearing that the FCC should have resolved this dispute before the auction.

“The situation we have is embarrassing,” she said. “We have to resolve these issues before we put the spectrum to market in an auction. We bring these airwaves to market, ask carriers to spend billions of dollars on them, and then don’t know exactly what the terms of service will look like.”

But Rosenworcel stopped short of saying the FCC’s current emission limits should be changed. “We have a bunch of different studies on the record, and I have not been in the meetings where we have gotten to the bottom of just what threshold for out-of-band emissions should apply,” she said.

https://arstechnica.com/?p=1522047