Fire trucks on the scene of a fatal crash yesterday near Montgomery County Airpark in Gaithersburg, MarylandReuters

Yesterday afternoon, after flying with my wife in a small propeller airplane up through the Central Valley of California (for today's Atlantic Navigate conference in San Francisco), we heard the terrible news that a small jet airplane had crashed into a house near the Gaithersburg airport in the northwest suburbs of DC, killing all three people on the plane and a mother and two children inside the house. This is a disaster on a smaller scale than an airline crash but in a way, more horrible, with the deaths of young family members as they went about their normal lives at home.

I am so sorry for everyone affected by this crash.

Only because I have some relevant local knowledge about this site—having flown in and out of Gaithersburg airport over the past 16 years and having based my small Cirrus airplane there for more than ten years—I am writing this post to add some basic facts.

1) Montgomery County Airpark in Gaithersburg, known as KGAI in aviation terminology, is a "small" airport but not small enough that its runway size or geographic position is likely to be a factor in this crash. Its runway is 4,200 feet long and 75 feet wide. For perspective, most big commercial airports have runways with lengths of 8,000 feet and longer. But small jets and turboprops go in and out of Gaithersburg all the time. On an average day, it has well over 100 takeoffs and landing. It is an active place.

2) The neighborhood where the crash occurred is very close to the airport, by national standards. This shot, from Google Earth, shows the place where the crash occurred, on Drop Forge Lane. The red line, which I've added, is the final approach course for Runway 14 at Gaithersburg, which the airplane would have been following.

Red arrow shows normal approach to Runway 14 at Gaithersburg. Blue arrow shows normal Runway 14 departure. Green arrow shows the recommended noise-reduction departure from Runway 32 at Gaithersburg (taking off on the same runway but in the opposite direction), to avoid flying over the neighborhood where the jet crashed yesterday.

The area of the crash was less than a mile from the runway threshold, again close by national standards. On a normal approach airplanes would be somewhere between 500 and 1000 feet above the ground at that point in their descent.

3) None of the subdivisions or commercial areas that now surround KGAI were there when the airport was built in the late 1950s. They have expanded as this part of the close-enough-to-be-commutable, far-enough-to-be-affordable part of the DC suburbs has grown. Technically, developers built there and purchasers bought there knowing an airport was nearby, but of course no one expects to have their home destroyed and their family killed by a plane.

4) Airport operations show an awareness of this neighborhood's concerns and existence. The "preferred calm wind" runway is this same Runway 14, so that planes would whenever possible take off headed away from this neighborhood—as shown by the blue arrow in the graphic above. When winds favor takeoff in the opposite direction—on Runway 32—pilots are supposed to turn to the right as soon as possible precisely to avoid flying over the neighborhood in question. That is what the green arrow shows. (Every runway is known by two names, depending on the direction the plane is going. The names are based on their compass heading and always differ by 180 degrees. If a runway goes straight east-west, planes headed to the east will be flying heading 90 degrees and using Runway 9. When taking off or landing in the opposite direction, they will be flying on a 270-degree heading and using Runway 27.)

5) The weather at the time of the accident was not perfect but was benign enough not to seem an obvious cause of a crash. The winds were light. The ceiling was around 3,000 feet—which means that the jet would have been flying under instrument flight rules through most of the flight and would have followed an instrument approach to the airport. (Rather than just picking it out visually.) But once it descended below 3,000 feet it would have had the runway in sight.

6) The location of the crash is not a usual site for either mechanical failures or the most familiar type of loss-of-control crashes near an airport. As a plane slows and descends for a landing, the main mechanical problems would be if the landing gear did not come down or the flaps (which allow the plane to keep flying at a slower speed as it prepares to land) did not extend. But the pilot would have been aware of those problems and would have mentioned them in radio transmissions, which he did not.

The usual loss-of-control accidents near an airport occur when a plane makes too tight a turn when flying the rectangular "traffic pattern" in preparation for landing. The FAA image at right shows the "base to final" turn before landing. If a pilot mismanages that turn, the plane can stall (lose lift) and fall to the ground.

But in this case, the jet would have been following an instrument approach (probably RNAV 14) to Gaithersburg, which would have given him essentially a 10-mile long straight-in approach to the runway, with no need for this last minute turn. Also, Runway 14 has a "VASI," a set of red and white lights to give an incoming pilot a guide to the proper descent slope to follow. If you see all red lights, you're too low; all white, you're too high. Both red and white, you're on the right path.

7) To summarize #5 and #6: None of the usual weather-related, mechanical, or traffic-pattern problems that explain crashes seem to apply in this case.

Update 7A) The recording of radio traffic on the KGAI frequency contains several references to large number of birds around the runway. In my experience that's not rare, but it is conceivable that a bird-strike could have disoriented, distracted, or even disabled the pilot; or that maneuvering to avoid birds could have led to a loss of control; or that a bird going into an engine could have been the beginning of the plane's problems. The recording is here. [Update-update: see NTSB briefing below, which finds no evidence of bird strike or "bird ingestion."]

8) Gaithersburg is an "uncontrolled" airport, with no control tower. As the recent midair collision near Washington showed, control towers don't eliminate all traffic-conflict problems. But at Gaithersburg, pilots judge their position relative to one another through announcements on the CB-style common radio frequency. "Montgomery County traffic, Cirrus XXX is nine miles to the northwest, will make 45-degree entry to right-downwind for Runway 32." "Montgomery traffic, Cessna XXX is turning base to final for Runway 32." Etc.

9) Gaithersburg is very active as a training airport. On good-weather days (and yesterday was good enough to qualify) its environs usually contain a number of planes doing takeoffs and landings as part of their drill. Often they fly "closed traffic": taking off, flying the rectangular traffic pattern, landing, and doing it again. By definition, many of these are less-experienced pilots. A lot of them are non-native speakers of English, which means that it can take them longer to report their position and plans on the frequency, and sometimes to be less accurate about it.

10) The combination of points 8 & 9 can complicate the final stages of approach to landing at KGAI, in the following way: If you're coming in on an instrument approach, from ten miles out you're on a straight line for Runway 14. But the closer you get, the more you're alert for student (or other) pilots taking off, landing, or flying around in the pattern. I've kept count, and in recent months on about half the approaches I've made, I've had to make close-to-the-airport adjustments because of traffic in the pattern or whose location I wasn't 100 percent sure of. Sometimes this meant "going around," putting in power, climbing, and circling around for another landing attempt. Some times it means slowing down or making delaying maneuvers, usually "S-turns" to draw out the arrival process and sometimes full 360-degree turns. It's an expected rather than startling aspect of operations at this airport.

On the probabilities, I can imagine something similar happening as the light jet neared the airport: traffic the pilot hadn't expected meaning he had to adjust his plans, and then something going wrong from that point onward. There is nothing inherent in a delaying turn that would make it dangerous—S-turns involve a shallower bank than right-angle turns in the traffic pattern. But obviously something, as yet unknown, made the pilot lose control of the plane. Traffic in this area is all carefully monitored, as part of the special security rules in the DC area. So it should become apparent which aircraft were where, and when.

11) As a matter of public record, the pilot of the plane, who with his two passengers was killed, had been involved in a different loss-of-control landing accident in a different airplane four years ago at the same Gaithersburg airport.

Sincere sympathies to all on this terrible event.


News update 6:20pm EST The Aviation Safety Network has relayed this additional information, which bears on some of the possibilities mentioned above. I'll just post this now and do further explanations later:

The following preliminary findings -all subject to be validated- were reported in an NTSB press briefing on December 9:

- Flight time from Chapel Hill to Gaithersburg: 57 minutes
- En route altitude: FL230 [approx 23,000 feet]
- Captain (ATPL rated) seated in left hand seat [ATPL=Air Transport Pilot, an advanced-proficiency rating]
- Passenger seated in right hand seat
- Flight was cleared for RNAV GPS runway 14 approach
- 46 Seconds before CVR [cockpit voice recorder] recording ended: Radio Altimeter callout of 500 feet
- 20 Seconds before CVR recording ended: Audio stall callout, which continued to the end of the recording
- Flaps were extended and gear was down
- Lowest recorded airspeed by FDR [flight data recorder]: 88 knots
- Large excursions in pitch and roll attitude were recorded by the FDR
- 2 Seconds after lowest airspeed was recorded, the throttles were advanced
- No evidence of engine fire or failure or bird ingestion

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.