Lesson From the "Miracle on the St. John's:" DON'T LAND IN THUNDERSTORMS!

I started this blog in November, 2009, and during that time I have periodically warned the aviation community of the continuing danger of thunderstorms. I have cited anecdotal evidence that the pilot community has lost their "fear" of thunderstorms since the downburst was conquered. I wrote this after the serious American Airlines excursion into a thunderstorm this past June.

Last night, at Jacksonville Naval Air Station south of downtown Jacksonville, a Boeing 737-800 ran off a runway and slid into the St. Johns River. Fortunately, all survived with only a few minor injuries. The crash occurred at 9:40pm EDT.

They are calling it the "Miracle on the St. Johns" in an allusion to the Miracle on the Hudson in 2009. According to the Jacksonville Times-Union (blue link above) the passengers were standing on the wings awaiting rescue as they did in NYC ten years ago.

The Hudson accident was caused by an unavoidable bird strike. The Jacksonville crash was, at least in part, caused by the decision of the flight crew to land in a thunderstorm.

Here is a preliminary rundown of the meteorology of the situation. Let's begin with the radar.

Above is the reflectivity data from the Jacksonville WSR-88D radar (18.8 mi. away) from 9:38pm, two minutes before the crash. The blue dot is JAX NAS. The -88D radar shows (and the onboard radar should have shown) very heavy rain falling at the time of the approach and crash. Indeed, the aviation surface weather observations (known as METARs) from JAX NAS show just that.

For those who are not pilots, here is a translation (courtesy NCAR) for the observations before and immediately after the accident.

Weather observations are in Universal Time. So, 9:40pm EDT = 01:40 UTC

At both 9:22 (0122 UTC) and 9:45 (0145 UTC), there was a thunderstorm in progress with "frequent lightning overhead" (from the aviation METAR; click on it). The winds were gusting to 18 mph immediately after the accident which the crew should have been able to handle.

There is no evidence of a downburst, at least so far. The surface winds from the METARs do not suggest it nor does the radar show evidence of one.

The Doppler wind data (above) shows the same thing as the surface observations: A northerly component to the wind but the winds are not particularly strong and no abrupt changes in speed or direction beyond what one would expect in a thunderstorm. We'll come back to that sentence.

If it wasn't a downburst, what might have caused the crash, at least from a meteorological perspective?

• Very heavy rain causing skidding or hydroplaning. This seems possible since the plane ran off the runway. 
• Poor visibility during the approach and landing.
• Shifting winds -- while not a downburst, winds frequently shift or vary in speed and directions in thunderstorms which is one of their inherent hazards. 
• A lightning strike. While the last commercial airline crash due to lightning was in 1967, a strike causes additional work for the flight crew. During landing, this could be a serious issue. 

The pilots should have seen the lightning from a considerable distance. The GOES-E weather satellite shows quite a few strikes in the immediate area (yellow, red, blue dots).

As the plane was enroute to JAX NAS from Guantanamo Cuba, it would have been approaching from the south or southeast. The above picture also shows low clouds (grays) with the isolated tall cumulonimbus cloud (the thunderstorm) over the air station. That means the storm would have been clearly visible to the crew while it was being illuminated by lightning.

These are some preliminary ideas. The NTSB is, fortunately, going to do a full investigation.

That stated, I cannot emphasize this enough: The exhortations to pilots not to takeoff, land, or fly through thunderstorms are there for a reason. When there is a thunderstorm over the field, either circle or divert. While pilots periodically get away with landing in thunderstorms, it is a form of Russian Roulette. Don't do it!