The National Weather Service’s warning dissemination system – critical to its mission of protecting life and property – failed today during the peak of a severe thunderstorm outbreak affecting the D.C. area and much of the Northeast.

This outage had local consequences.

At 4:12 p.m., a severe thunderstorm warning for golf ball-size hail and winds over 60 mph was issued for D.C.’s  eastern suburbs until 5 p.m.  It included Odenton, Crofton and Annapolis. The warning never reached the public through normal means: on NWS Web sites, mobile devices, electronic relays to broadcast media (via commercial weather providers and text messaging systems), etc.

In providing updates for readers of the Capital Weather Gang, the warning never made it to me (I subscribe to NWS text alerts and also ordinarily get warnings through a Twitter service).

Online radar imagery (on the web and smartphones) was also unavailable during the outage outside the NWS, significantly impairing the public’s and media’s ability to track storms.

An email obtained by the Capital Weather Gang – with the subject line “Major Disruption to NWS Product Issuance” (see below) – reveals systems went down between approximately 4:06 and 4:37 p.m. ET.

“It appears that all NWS warnings did not properly disseminate during the outage, and significant severe weather was occurring during the outage,” the email says.

Some local forecast offices, including the  office serving the D.C. area in Sterling, Va. scrambled to social media platforms in an attempt to manually get warnings out but their reach was severely limited.

Fortunately, when the NWS issued a tornado warning for central Anne Arundel County at 4:38 p.m., systems had just come back online.

The NWS email says “a firewall upgrade” caused the problem.

Still gathering the facts but what we know so far … it appears a firewall upgrade caused the anomaly lasting 31 minutes. We are looking into the extent in which product issuance may have been affected. Product issuance seems to be back to normal now.

This is not the first time National Weather Service’s computer systems have failed.  Almost exactly a year ago to this date (May 23, 2013), I published a piece headlined: “Weather Service systems crumbling as extreme weather escalates“. Excerpt:

In the past 5 days alone, a telecommunications outage near Chicago made it difficult for NWS forecasters to issue warnings, a major weather satellite failed, the website for the entire NWS Southern Region went down, and a NWS official in tornado alley declined to launch a weather balloon citing budget concerns.

Slate’s Eric Holthaus summarizes some more recent computer failures just in the  past six weeks:

Previous technicalities weren’t quite as serious, such as an unbelievably large flood warning in mid-April (coincidentally, the day before the movie Noah was released), and a website crash in early April (though warning products continued to be issued as normal through more traditional channels).

Weather forecasters around the country, who rely on NWS radar, forecast products, and warnings to inform the public of weather hazards, took to Twitter to rightly criticize this outage.

Slate’s Holthaus offered this commentary:

Outages like this can’t continue. The National Weather Service should immediately implement redundancy into their computer systems to ensure the people they serve—us—aren’t kept in the dark when dangerous weather approaches. Until then, the system in place is an embarrassment to the diligent scientists that work there.

I agree with Holthaus.  The National Weather Service can have the best forecasters, the best models, and the best radar in the world.  But if it can’t get its information out, it is failing the public.