Weather Satellite Data Hacking and Breakdown: Why It Matters for Forecasting

Placeholder while loading article actions

NOAA’s satellite data and information service network was hacked in September and, in response, a major weather forecast data stream was taken out of service for two days in October, while NOAA was performing “unscheduled maintenance” to fix the problem, sources said. .

The maintenance, due to its relatively short duration, had little impact on weather forecasts, but the hack revealed a vulnerability that could have significant implications for weather forecasts if the satellites or satellite network from NOAA were again compromised.

We released the full story of the satellite outage and hacking incident early this afternoon. An extract:

“NOAA told me it was a hack and it was China,” said Wolf, who also chastised the agency for not disclosing the attack “and deliberately misleading the American public into error in his answers.

“They had an obligation to tell the truth,” said Rep. Frank Wolf, (R-Virginia). “They covered it.”

The story highlights the importance of satellite data for weather forecasting, but I can offer a little more context here.

NOAA leadership, in public statements, has repeatedly spoken of the value of satellite data as an essential component of the nation’s “environmental intelligence.”

“NOAA satellites provide critical data for forecasts and warnings that are vital to every citizen and to our economy as a whole,” NOAA Administrator Kathryn Sullivan said in a 2013 statement.

After recent high-impact weather events, NOAA conducted experiments demonstrating the critical value of satellite data.

In 2012, NOAA said its satellite data was instrumental in the European Center for Medium-Range Forecast (ECWMF) model’s 5-day prediction that Super Hurricane Sandy would spin and hit the northeast coast – where it caused more than $50 billion in damage – rather than hanging out to sea. “If we had thought the bulk of the storm was going to stay in the Atlantic, or if the locals had only had a day to prepare or evacuate, the results would have been even more devastating,” Sullivan said at the time.

Related: Without polar satellites, forecasts for super hurricane Sandy would have suffered from the findings of European analyzes

In 2011, NOAA credited satellite data for his ability to successfully alert the public 5 days in advance of the likelihood of a dangerous severe weather outbreak, which eventually spawned 312 tornadoes from April 25–28. “We would not have been able to alert the public as soon as we did to the high potential for these violent tornadoes, without NOAA satellite data,” said NWS forecaster Greg Carbin. .

And in 2010, the NWS said satellite data proved essential in its forecast for “Snowmageddon,” the February blizzard that paralyzed Washington, DC and much of the mid-Atlantic. The NWS said five-day forecasts made without the benefit of satellite data would have underestimated snowfall amounts by at least ten inches in the DC area.

“Our global observing systems are the foundation of the information and data we provide – without them we are ‘flying blind’ and the level of uncertainty in our forecasts increases at a time when people are demanding more precision”, Sullivan testified before Congress in April.

The October satellite data outage, precipitated by the hack, meant that weather forecasting centers at the National Weather Service and around the world did not receive large amounts of data from NOAA’s weather satellites.

“All operational data sent through NOAA, which is normally an excellent service, has been lost,” said Stephen English, head of the satellite section of the ECMWF, which is renowned for running the world’s most popular weather forecasting model. advanced and most accurate in the world.

NOAA said the two-day outage in October negatively affected the “competence” or accuracy of its prediction models, but probably not enough for the public to notice. According to NOAA, the outage caused the model’s forecast to degrade by about 12 hours, so its 5-day forecast was about as accurate as the typical 5.5-day forecast.

ECMWF English said their model was “resilient” to failure and did not experience a noticeable drop in accuracy. “Had the outage lasted longer, the impact could have become significant,” English said.

The outage also affected Environment Canada, which operates a global prediction model. The impact of lost data, he said, was difficult to assess.

The October data outage, in addition to exposing security vulnerabilities in NOAA’s satellite network, may in some ways be a preview of the predicament NOAA may face beginning in 2016 due to of a possible gap in the coverage of satellites in polar orbit. The gap is believed to result from a rift between the design life of one of its current satellites and the planned launch of its replacement in 2017. In 2013, the Government Accountability Office cited the impending gap as one of the top 30 challenges facing the federal government. .


Comments are closed.