Walden Probes What Went Wrong in National EAS Test

Audio was a major problem encountered during the test, which overall was a success, FEMA Assistant Administrator Damon Penn told a separate hearing on Capitol Hill. Democratic and Republican members of the Homeland Security’s Committee’s Emergency Preparedness Subcommittee had questions for Penn on the test. He said the agency ought to be able to provide initial data assessments soon, and it will take a few months to put together a more comprehensive report. “I think the test was comprehensive enough to give us a good start” on things to fix, he testified: “We have plenty” to address now, and FEMA wants a future test to last two or three minutes. The simulation was initially to have lasted three minutes, and was reduced to 30 seconds a few days beforehand (CD Nov 7 p6).

"What they've proven is: it sort of works,” Walden said. “That’s not good enough.” He called for another national test after the agencies determine what went wrong the first time. Walden said he had no concerns with the agencies’ process behind the test. “I actually commend the FCC for initiating the test and FEMA’s role in that,” Walden said. “We needed to find out if it would work, and it didn’t.” He praised FEMA’s progress deploying more primary entry point (PEP) stations and developing a redundant satellite capability.

Sixty of 63 U.S. PEP stations functioned correctly and the agencies are “trying to figure out what went wrong” with the other three, Walden said. The FCC heard back from about 60 percent of broadcast stations, and learned that 78 percent of those received and retransmitted the test, he said. “In some cases, they think that the header that gets sent may have opened and closed the relays at the stations, which would explain why there was just an audio burst and then no audio after that,” Walden said. “And in other cases it didn’t get through at all, and there may be some proprietary issues in some of the equipment.”

Walden was in the control room at KCBS Los Angeles during the test, and witnessed the problems firsthand, he said. He went to KABC afterward, he said. All KCBS got of the message was, “This is a,” he said. KABC “didn’t trigger automatically,” he said. “They had to trigger it and then they were able to play a certain percent of the audio.” While Walden doesn’t think a hearing is necessary, he asked FEMA and the FCC to “get back to us with their evaluations of what transpired.” Walden didn’t give the agencies a deadline, he said. The House Commerce Committee will make that information public, he said.

FEMA appeared to be part of the reason why the test couldn’t be heard well by some radio listeners and TV and subscription-video viewers nationwide, Texas Association of Broadcasters President Ann Arnold told us. She cited her conversations with government and industry officials who participated in the test. FEMA apparently programmed the test to start a few minutes after 2 p.m. EST on Nov. 9, which might have contributed to some audio problems, Arnold said. A phone-line link from the agency to PEP stations, which pass on warning messages to other EAS participants, might also been a problem, although it’s unclear if the agency or the PEP stations caused it, she said.

The test began a few minutes late because FEMA programmed the emergency alert notification code, which alerts EAS participants’ encoder-decoders that there’s a message coming, to start with a slight delay, Arnold said. She recounted that several companies that sell encoder-decoders to EAS participants said the EAN was programmed to not be valid until a few minutes after the scheduled start. Audio during the test was garbled in some instances because the return-audio link of a phone patch from FEMA to PEP stations wasn’t muted, Arnold noted. The exercise overall was a success, because all or almost all EAS participants in Texas received the test, she said. “Even though the audio was garbled,” there “was enough to let people know that something was going on,” Arnold said: That’s “more than I expected."

A PEP station’s encoder-decoder malfunction caused audio feedback to be rebroadcast back into parts of the system, causing the message to be “garbled,” Penn testified. That was “a large part of the problem,” he said. “Audio quality ... throughout the test was sporadic and in some cases didn’t exist at all,” he said earlier at the hearing. The message on the screen didn’t always match the audio, which “needs to be better, and we can do that,” Penn said. “We need to do a much better job before we use this scroll.” He said there were “mixed reports” from broadcasters, DBS providers and cable operators (CD Nov 10 p2) that had “specific issues that we need to work through.”

Ranking Member Laura Richardson, D-Calif., thought Penn “gave a fair assessment of the national test and what occurred,” she said. Chairman Gus Bilirakis, R-Fla., had asked at the hearing’s start for the various witnesses from the Homeland Security Department and its offices to discuss “any successes and gaps” identified by the test. Witness statements are at http://xrl.us/bmivtv.

Just that the test occurred “was a success,” Penn testified. There was “some equipment we've had for 50 years that we never turned on” before the first-of-its-type exercise, he said. “Our message propagation worked better than we thought it would,” from the government to broadcasters, with in some states 90 percent of the public covered, Penn said. “The public was not overly alarmed that we were doing a nationwide test.” Some government and industry EAS participants had feared (CD Oct 13 p10) that people could think it wasn’t a test and panic or call 911, because in some cases the video didn’t clearly say it was a simulation. Since such panic didn’t occur, Penn said, “we owe that a lot to the broadcasters and the public service messages they put out and the backdrops during the test” which clearly stated it was an exercise.

Some public, educational and governmental channels didn’t get the test from cable operators, which sometimes pass on EAS warnings to them, said the head of a PEG industry group. There’s no requirement that cable operators do so, and in the past they haven’t always done it, said Alliance for Community Media Executive Director Sylvia Strobel. “Sometimes they've voluntarily passed it on,” she said of cable operators, and the test “did go through to some PEGs,” she told us. There are also “some communities where it didn’t go through for anybody,” whether a public access channel or a radio or TV station, Strobel said. PEG networks aren’t “necessarily required to do what the broadcasters do,” by sending alerts to viewers, “but we do see this as an important function of our members from an emergency management standpoint,” Strobel said.

The FCC should include PEG in its review of what during the simulation worked well and what didn’t, Strobel wrote Chairman Julius Genachowski Wednesday. She requested “an examination of whether viewers of PEG channels were provided with the alert.” Many of the channels weren’t “because the operators of many cable systems on which PEG channels are carried -- primarily large cable companies -- could not or chose not to participate in the test,” said the letter in docket 04-296 (http://xrl.us/bmivm4). Spokespeople for the American Cable Association, NCTA and Media Bureau had no comment. NATOA is surveying its members and those of the PEG group to see who got the alert, Strobel told us.