I was saddened to learn today of a Cirrus SR20 that crashed into a Manhattan skyscraper, killing Cory Lidle, the New York Yankees pitcher, and Tyler Stanger, a 26-year-old flight instructor from California. As it happens, I was at the exact same spot yesterday, en route to the Downtown/Wall Street heliport in a Robinson R44. I have flown my own Cirrus SR20 all the way to the Arctic Ocean and then southwest to Alaska. Friends have thus been asking me to speculate on how this accident might have occurred.
The airspace around New York City is extremely complex (chart). There are three Class B airports, LGA, JFK, and EWR, and therefore a lot of protection for jets flying in and out of these busy airports. You can’t be anywhere near the city unless you are talking to New York Approach, the tower controllers for one of those airports, or over the Hudson River between the surface and 1100′ above sea level (i.e., pretty darn low).
I have often been confused about the East River. Careful study of the helicopter and terminal area charts has led me to conclude, mistakenly, that it is not legal to fly too far north of the Brooklyn Bridge without explicit clearance from LaGuardia Tower on 126.05. As Lidle and his instructor probably knew from talking to local experts, it is in fact legal to continue up to Roosevelt Island, at 86th St., without talking to LGA. In any case, if you continue straight ahead up the river indefinitely without getting clearance (which is easy to get), you are very seriously violating LGA’s airspace. If discovered, the FAA could have suspended the pilot or instructor certificates of anyone on board the aircraft.
My preliminary best guess (and at this point it can only be a guess) is that the two pilots on board the accident SR20 were cruising slowly up the East River. At some point, they decided that they’d reached the end of the little cut-out tongue of uncontrolled airspace over the East River. They attempted a 180-degree turn in an attempt to get southbound down the river toward uncontrolled airspace. An airplane in a sharp turn stalls at a much higher airspeed than when straight and level. Merely by putting the airplane into a steep bank and trying to hold altitude, they could have gone from flying to an aerodynamic stall (wings at too high an angle to the relative wind or, in simpler terms, air not moving fast enough over the wings) in a matter of seconds. At this point, the airplane is not easily controlled and a lot of bad things can happen. Low-speed low-level maneuvering, which typically happens when aircraft are trying to land, is the leading cause of plane crashes.
[It is possible to turn an airplane tightly and safely and is commonly done inside mountain valleys in Alaska (where guys just love to take off and head towards a pass to see if there is any separation between the clouds and the terrain; if not, they turn around and go back to their cabin). The trick is to slow down as much as possible. An ice skater going fast will use up a lot more ice in a 180-degree turn than an ice skater going very slowly. In an airplane, this means putting out flaps so that you can fly slower without stalling and slowing down to maybe 1.5 times stalling speed (in the Cirrus SR20 this would be about 75 knots with two people on board). At a slow speed, you have to be somewhat careful with bank angle because you are closer to the stalling speed. On the other hand, you don’t need a steep bank angle to make a tight turn because you’re only going about as fast as a car.]
[Thursday update: I was interviewed by a New York City radio station this morning. The interviewer, as have most journalists, seemed very interested in the Cirrus’s parachute. People can’t shake the idea that a plane with a parachute is safer than a plane without one, though in this situation, the safest plane would have been an old slow cheap one that could be flown slowly and therefore turned tightly. The Cirrus is a great plane for going straight and level on a 400-mile trip, but its virtues become liabilities when trying to fly low and slow.
http://email@example.com is a video clip from the local TV news station that flew with me and Alex this afternoon]
[Saturday update: found a good table of airspeed/bank versus turn radius: http://selair.selkirk.bc.ca/aerodynamics1/Lift/Min_Radius.html ; my example above of slowing the Cirrus down to 75 knots would enable a turn diameter of 1000′ at a comfortable, yet steep-by-Private-pilot-standards, 45-degree bank angle. The East River is approximately 2000′ wide. At a more standard slow pace for a Cirrus of 100 knots and 30-degree bank angle, the turn diameter is 3075′.]
28 thoughts on “Cirrus SR20 crash in Manhattan”
One of the early reports said it was an instructor in there with him. Didn’t the Yankees’ season just end? I wonder if he’d been out of the airplane for a few months and decided to hire an instructor to get back up to speed.
The article I read said he had a cross-country trip planned, that his contract specificially mentioned the possibility a flying death, and that he’d talked up “the parachute” of the cirrus.
You’d think if they did a 180, they would have been inclined to turn right, away from the city.
What is that inclusion into the Laguardia Airpace? There’s a “tongue” that follows the river up to the north end of that skinny island (up to the left of where it says “UC”) According to google earth, they were halfway up that skinny, long island.
Just curious — you didn’t mention the reported mayday re: fuel problem. Why not?
The utter shock I am hearing from most folks not familiar with general aviation is that they had no idea private aircraft are even allowed near a large city like New York after the 9-11 attacks.
The general public seems to feel like general aviation shouldn’t be allowed near large metro areas anymore. I suppose this accident will fuel that opinion in many circles.
On another front, if the preliminary reports are correct, a CFI was on board. Wouldn’t (or maybe shouldn’t) a CFI be well-versed in such a manuver as you described, Phil?
Is it also true that although these model planes are equipped with parachutes that due to the low altitude they were flying the chute would have been of little help in the event this was an engine failure?
Mike: I wish that I could answer your question about LGA airspace, but honestly I have looked at that VFR terminal area chart quite a few times and can’t tell you what the legal altitudes over the East River are. From surfing around the Web and hearing an FAA spokesperson comment, I have become convinced that there is some sort of corridor most of the way up towards the middle of Central Park (i.e., about where they were maneuvering).
Splashman: I did hear something about a fuel problem, but (1) it seems unlikely that they were on LGA Tower frequency if they weren’t talking to LGA Tower; they should have been on the East River common traffic frequency (pilot-to-pilot), (2) you aren’t going to get bad fuel at Teterboro, (3) with only two people on board, you would probably top off an SR20 and have five hours of fuel, (4) with a CFI on board, I don’t think they could have overlooked a nearly empty set of tanks, (5) if you ran out of gas or your engine quit, you would land straight ahead into the river or on a highway, not start doing steep turns (it is legal to enter LGA airspace to make an emergency landing!).
Mark: Shouldn’t a CFI be able to do a steep turn? Sure, but remember that most CFIs are fairly young with low time and they are doing the job to build time. Remember further that CFIs have most of their experience in forgiving trainer airplanes such as the Cessna 172. The Cirrus is a different beast and the lack of feedback from the spring-loaded flight controls makes it hard to tell that you’re going too slowly to sustain flight. Many studies have found that time in type (experience with a particular kind of plane) is more important than total time. Only a CFI who works for a Cirrus training operation would have a significant amount of time in the Cirrus (or a guy like me who owns a Cirrus and happens also to be a CFI, but at $14/hour it takes a long time to save up $280,000 for a new Cirrus!)
Mark (again): I think the Cirrus parachute deployments are supposed to be somewhat effective even as low as 400′ above the ground. A water landing in the East River would have been preferable to a parachute landing, I think. The Cirrus parachute landing into water that I heard about resulted in serious injury to the pilot. Apparently you need the landing gear to strike something solid and absorb a lot of the force. Something like 85 percent of ditchings (landings in water) are survived and that includes guys who land in the open ocean. A plane will float for a few minutes after landing (unlike a helicopter, which sinks like a rock).
This is all speculation, of course, but I think the most common cause of accidents is the most likely: a botched sharp turn close to the ground.
[One ironic fact that will probably be overlooked is that if they had been in a 30-year-old $18,000 deathtrap of a Cessna 152 instead of their modern super-safe parachute-equipped $200,000 Cirrus, they would be relaxing over a beer right now. The 152 flies a lot slower, so it would naturally turn tighter without a steep bank. The 152 would give more warning in the controls before a stall. The CFI would probably have had a lot more experience with the 152.]
This is interesting:
I also wonder if it’s possible his avionics told him he was violating some kind of airspace restriction and he paniced and tried to turn around going too fast too low.
The distress call is very odd. If the plane had stalled, would the pilot have had the time (or the mental reserve) to transmit a mayday?
I fly helicopters and have not a single hour in fixed-wings but wouldn’t a stalled plane come down really fast?
For a pilot to have the wits about him/her to transmit a distress call seem to me to indicate a malfunction in the plane itself, maybe?
Surely they didn’t get into IMC (instrument meteorological conditions) did they?
Heh. No sooner do I read reports (CNN, etc.) of a fuel-related mayday and post my question here, then I read that the Feds are backing off (http://abclocal.go.com/kabc/story?section=nation_world&id=4652429) on their claim of a mayday. So, Phil, it appears your skepticism was on-target. Kudos!
As a CFI in a Cirrus, what do you think about the idea that the CFI could actually take over the control of the aircraft if an athlete has a firm panic grip on his side stick control? He was right hander but still in good condition.
Quote from http://netlib.bell-labs.com/who/sape/pilotage/Hudson/index.html. Good map on that site.
“After the Lady, we swung around Governor’s Island into the East River. The East River has its own self-announce frequency of 123.075. Do I have to tell you to get an up-to-date VFR Terminal Map for New York?
At the north end of Roosevelt Island, the East River dead-ends onto LGA’s air space. I always turn around well south of the Island, where the river is nice and wide. I warn the passengers of the G-forces and usually make a 60°-banked turn (it’s my sadistic streak – 45° is plenty to make the turn). Watch the wind – it’s usually from the west, so a left turn is into the wind. Rarely, winds are from the east, however, and then a right turn may make more sense to keep the radius small. Make sure you announce well and look behind you before turning. Do I need to tell you to be proficient in steep turns before venturing into the East River? This is not a place to practice them.”
By the way there was an east wind all afternoon so a right turn would have been the proper choice, not the apparent left turn
4 PM (20) Oct 11 62 (17) 57 (14) 29.85 (1010) E 13 rain
3 PM (19) Oct 11 64.9 (18.3) 55.0 (12.8) 29.88 (1011) E 15
2 PM (18) Oct 11 64 (18) 55 (13) 29.91 (1012) E 13
1 PM (17) Oct 11 64.0 (17.8) 55.0 (12.8) 29.93 (1013) E 10
Noon (16) Oct 11 63.0 (17.2) 54.0 (12.2) 29.95 (1014) E 12
The best chart is the NY Helicopter Route Chart, Downtown Manhattan Inset. There is an East River VFR corridor from the southern tip of Manhattan up to the North tip of Roosevelt Island. The chart indicated controlled airspace extends from above 1100 feet up to 7000 feet (the top of the NY Class B).
The general rule is that you treat the corridors like roads, keeping to the right. I have flown them too many times to count and while busy, they are nowhere as busy and hectic as flying into Oshkosh. You turn on every light on the plane, monitor and announce position on the proper frequencies (123.075 for the East River), and keep your head on a swivel watching for other aircraft, especially helicopters typically at 500 feet. I typically will fly at 1000 feet while VFR but if it looks really busy I will frequently contact NY Approach and request entry into the Class Bravo and do the same route at 1500 or 2000. LGA tower is *very* accomodating, by the way, to requests for the Bronx or Throggs route.
I am still mystified as to how this could have happened. As a pure guess, I might say that they were in a left turn at the top of Roosevelt Island, with a 15 knot wind behind them, and the turn was just too wide. I have made that turn a few times and didn’t find it that bad, but I was in a slower plane (Grumman Cheetah) and always turned into the wind. Even if I had to turn to the right. I just announced what I was doing and made sure there was nobody in the way.
But, that building is hard to miss! Did they lose control? Was there a mechanical failure? At the time the ceiling was 1800 feet, and a Coast Guard video from the North looking South showed the entire building was visible, even from several miles away, at the time of impact.
As a side comment, the number of totally brain-dead, stupid things said by supposed “experts” interviewed by the media is astounding.
Mike: the VFR corridor runs exactly to the northern tip of Roosevelt Island. You can go that far, but no further (without a clearance from ATC).
Mike, Tom: I’ve flown the Hudson but not the East River. But the East River has the same keep-to-the-right convention, doesn’t it? So perhaps they made a left turn for that reason, despite the east wind.
Philip: You may be right about overbanking and stalling, but I’d be more inclined to guess the opposite–that they tried too hard to avoid overbanking and stalling. As a result, they were too fast and turned too slowly, making the turn too wide (until perhaps the last moment, when they may have turned steeply and stalled while trying to avoid the looming buildings). If the turn had initially been too sharp instead of too wide, they’d probably have crashed into the river or island rather than into Manhattan, wouldn’t they?
It appears from this article from ESPN Host Allan Schwarz (http://sports.espn.go.com/mlb/columns/story?columnist=schwarz_alan&id=2622245) that Cory Lidle was familiar and comfortable enough flying over the East River to offer Mr. Schwarz three weeks ago a ride over the East River in Lidle’s Cirrus SR20.
The quote from Cory Lidle to Allan Schwarz (as reported by Mr. Schwarz): “you should really come up with me. We can fly right past your apartment building. You’ve never seen Manhattan ’til you’ve flown right up the East River. It’s beautiful. We can do it one day before a game.”
This is probably old hat to most of you, but I thought it was pretty cool: This is a Google Earth view over the river, looking North past the building that was hit. The cool bit comes in when you use the 3D airspace polygons Google Earth overlay from fboweb.com. I don’t know how accurate this is for “seeing” the airspace boundaries in that area, but it might help visualize in a general sense.
I have one theory about the crash:
The entire building could have been in a blind spot: directly below or above the aircraft when banked for the turn. Couple that with too wide of a turn and a wind blowing towards the building: suddenly the building comes into view, and you’re headed right for it!
I came within feet of another plane in the pattern one time when I was doing touch and go’s: The tower told me to follow a blue and white Cessna turning downwind. I saw a blue and white Cessna on downwind, and thought that was it. I turned downwind and followed it. Turns out the Cessna I was SUPPOSED to follow had a student pilot and an instructor in it, and was flying really low and slow, beneath me! I crossed over about 15 feet directly above him when I made my downwind turn!
The weird thing is nobody knew about it except 2 people on the ground who happened to see it! Not the tower, not me, the instructor, or the student pilot!
Every plane has a blind spot, and something big can hide in it at the right angle.
One thought to support the “blind spot” theory… 2 points.
1. If any issues come up in with the MFD (airspace warning, fuel, etc…), we tend to keep eyes down inside- looking at airspace, or whatever.
2. The building they struck seems to be a bit of a stand-out. The other buildings, in the immediate area and to the north, appear to be much lower. It certainly seems likely that with other things going on, they simply did not see that building until it was too late.
The minute I heard about this crash, I knew it was the same type of plane my brother flies. What a tragedy, with Lidle leaving behind a 6 year-old child.
Philip’s four-year-old nephew loved the helicopter ride, but I can’t say I’m going up in one of those again unless I’m shipwrecked or similar. I’m glad Philip has an engineering-type degree — somehow gave me confidence in his flying abilities when we did the “city tour” in the Robinson.
I’m guessing if Philip ever gets married and has any kids that his days as a flight instructor will be numbered. Is the median age of flight instructors extremely young simply because very few women have the nerves to worry about that on a daily basis? I noticed the instructor with Lidle was 26ish.
I’ve read all the comments,but after reading your pilot description of the Cirrus sr20 and all of its’ plus factors and as well minus factors, I think you have hit the nail on the head that it was pilot error and the pilot was the CFI who obviously did not have access to enough flying time in that plane. I like one writers comment as to the many stupid reasons given by experts.
Tom — unlikely. Any good CFI knows that all you have to do to get any student to let go of the stick is cover their eyes. Works every time.
Suzanne – the reason that CFIs tend to be young is that as soon as they accumulate enough hours to get a commercial piloting job they’re gone.
doesn’t the Cirrus SR20 have an onboard audio alert system – would they not have heard a verbal warning telling them they are about to hit terrain or would that warning have come too late to get out of the jam; or could it have caused them to panic ultimately hastening their demise?
Robin: They were flying an older Cirrus, which probably wouldn’t have had a terrain warning system (the new ones usually do, though our 2005 does not because we are cheapskates). In any case, they would probably have had to disable any terrain warning system during a flight that low near a city because it would have been squawking constantly. We have a handheld Garmin 296 in our Robinson R44 helicopter and, even on the lowest sensitivity setting, it gets unhappy when we fly around the skyscrapers of downtown Boston. We don’t have a cable running from it to our audio panel, so we only get on-screen text and graphical warnings.
The terrain warning system is most useful for (1) flying low in a helicopter where you might miss a radio tower, and (2) flying in the clouds in an airplane where the terrain warning system provides a backup in case you get grossly confused about your position on an instrument approach and are heading toward a mountain. When you are intentionally maneuvering close to obstacles, I don’t think the system is useful.
Suzanne (sis): Flight instruction is actually safer, statistically, than flying cheap little airplanes for transportation. If the weather is miserable and you’re stuck in a motel in Winslow, Arizona trying to get to California, you are tempted to take off. If the weather is miserable and you’re preparing to leave your house in Cambridge and drive out to Bedford to teach a guy how to fly around in circles over Hanscom Air Force Base, you call the guy to reschedule.
This was briefly mentioned, but again quickly seemed to go out of the minds eye.
What would really happen if a private general aviation plane had to be intercepted by a military class jet? Assuming even with a Mach 1.8 top speed that it actually got to the area before the plane was to be intentionally or unintentionally crashed.
Can a military Jet easily identify a plane such as this? At what range? What does a sidewinder do to a small plane such as this, and can it even engage it?
The closest situation I can think of was Pain Stewarts Jet that was flying on autopilot and gave the intercepting jets enough time to engage it if necessary. But given this and that kid landing at Red Square, are these planes too small and numerous to effectively track?
Anyone have any ideas?
“People can’t shake the idea that a plane with a parachute is safer than a plane without one, though in this situation, the safest plane would have been an old slow cheap one that could be flown slowly and therefore turned tightly.”
Well stated. As someone with only a little flight training, but most of my flight time in taildraggers in Alaska, this sentence really jumped out at me. I can’t fathom flying in NYC’s airspace… I’m used to flying in and out of airports where there’s not even a tower and landing on gravel strips where the main concern is whether or not there’ll be wildlife in the way. If I could have any plane of my choice, I’d probably ask for a shiny new Husky, but I’d be absolutely delighted with a trusty old Super Cub. Bet you don’t see too many of those flying around New York, though (I visit frequently, was just there last month… next time I’ll keep an eye on the skies a bit more to see what people are flying).
Jared: Watch United 93, which, I think was a fairly accurate reconstruction, to see what our military capabilities are like. Due to a combination of factors, not least the fact that all commands go through a large bureaucracy and mostly by telephone, it takes our military a long time to get ramped up. When the military is already tasked with intercepting small planes and in the area, they seem to do a fine job. I was flying the R44 into PSM back in August and heard a Piper Malibu calling air traffic controllers to ask about the F16 that was shooting flares at him. Tellingly, the controller had no idea who or where the Malibu was and had to go through the standard process of “type in the following transponder code and then squawk ident” then had to go offline to make a phone call or two and figure out how to coordinate with the Air Force. But that was a case where they knew in advance that they were going to have to do a few interceptions over a few days and just kept F16s circling constantly (we strengthen America against our oil-rich enemies by buying more Jet-A fuel from those enemies… :-))
Jacqueline: You see plenty of Piper Cub-style planes in the Northeast! All of the banner towers are in slow old planes like this. The main difference is that up here we don’t carry a big file in our aircraft to smooth out dings in the prop after landing on a gravel strip.
I don’t think it matters much if the military can intercept a small general-aviation craft — they simply aren’t capable of doing much damage, and the possible collateral damage from trying to shoot them down may be worse. Would you rather have a Cessna crash into your office building or a Sidewinder missile?
Even loaded up with explosives, a small plane has a miniscule cargo capacity compared to an automobile. The real urban threat is a truck loaded with conventional explosves (ala OK City or the original WTC bombing) or a dirty bomb. Small planes are irrelevant.
Below is a link to an Associated Press article (via ESPN.COM) with news that the NTSB cites pilot error and wind as the probable cause of Lidle’s fatal plane crash. It’s a preliminary ruling; the final ruling will be at a later date.
The report states that Lidle’s plane attempted a U-turn with only 1,300 feet of room for the turn, which would have required the aircraft to bank so steeply that it might have stalled according to the NTSB.
The thoughts of an old CFI.
I can’t imagine a CFI of any experience allowing his student to attempt a 180 down wind turn at below the roof top levels, in the confines of the East River. Only a very bold pilot (or a very stupid one) would intentionally attempt such a maneuver. One only needs to remember his S turns across the highway or turns about a point as a student to realize the effect of wind in a down wind turn.
With that in mind, I give Lidle’s CFI–Tyler Stranger, the benifit of expert judgement. I do not believe that he would even think of making such a turn at that altitude, while being aware of his proximity to the buildings.
I strongly feel that an event of catastrophic preportion occurred. A sea gull through the windshiel, for example. The airplane would be instantly out of control.
I have never flown the corridor. But I have crossed the river by ferry in days gone by and seen the gulls that trial the pleasure boats, searching for the bits of garbage that might be tossed overboard. I am sure that they still frequent the river, because that is what sea gulls do. And they fly high, too. I have dodged them at several thousand feet over the New England coastal areas, often far inland from the sea.
The NTSB and the FAA made no mention of finding any evidence of a bird strike. Obviously, evidence such as feathers and flesh would have been destroyed in the resulting fire.
Regarding turning in a small radius:
In the case of Lidle, it should have been obvious that a turn is made into the wind if the terrain is reasonably flat as in the NY corridor. Pilot inexperience and error.
I fly in Colorado.
In the case of flying up a mountain valley, which usually narrows as you approach the vertex, and you are obviously not going to make it (airplanes climb at 7 degrees, mountains rise at 35 degrees or more; the general angle of repose) you are left with little choice if you wait too long; there is not enough space to execute level turn. There is, however, a manouver that can save the day: use a hammer head turn. If practiced out on the flat it is an easy solution to that problem. The features are that, even in best angle of climb speed you can trade that speed for enough altitude to execute a hammerhead and come back down nearly on your same track and recover above the altitude you started at. The execution is: assuming that you are already at full power, best angle of climb and realize that you won’t clear the pass, pull the airplane up in as fast a “launch” climb as you can, trading speed for altitude, then just a you begin to lose alieron control, kick left rudder all the way (single engine plane), the airplane will turn about the yaw axis and head down in a nearly vertical attitude. You keep the yoke back as far as possible as soon as the nose is pointed down and you will recover straight and level a little higher than your entry altitude and heading 180 degrees from your entry course. Yes it takes practice but it can be learned and used with confidence. Go out to a practice area, find a road, get about 4000′ AGL and try it. Start slow, approach at a near-stall, power off, pull the nose up as hard as you can, kick the rudder, spin about and pull back the yoke. Progress to faster speed and more power until you feel confident to do it at full power and best angle of climb (the position you are most liable to find your self in when realizing that a) you won’t make the pass and b) there isn’t enough room to level turn).
This is all educated speculation. They weren’t flying below the building tops. It is almost certain the airplane became stalled while making the turn at an altitude above the buildings. While the physics of the turn show that it can be done with, what, a 45 degree bank, nobody does that kind of math prior to a flight. Pilots turn in a way that feels natural. Most likely they started with a 20-30 degree bank turn. Things looked to be going well at first, then it became apparent they would have to steepen the bank. To compensate for not being at 45 degrees the whole time, the second half of the turn would have had to be much steeper. Given that this is speculation and I’m not going to do the math, let’s say it needed to be 60-65 degrees of bank to finish the turn (most likely with full rudder and opposite aileron to stop the overbanking tendency). At that bank angle, stall speed for the Cirrus drastically increases (and despite all the comforting self-press from Cirrus about their stall characteristics, these airplanes are actually very finicky in a stall, especially with wings not level). They most certainly entered a stall/spin and descended in uncontrolled or semi-controlled flight. This would have been from an altitude too low for the parachute to be of any benefit. At that low of an altitude, traditional stall recovery of pitching down and adding power would have flown them right into a building anyway. This is by no means the first accident of its kind, even in a Cirrus. It is only high-profile because it involved an athletic star and a building in New York City. This accident has happened many times on a base to final turn, with non-famous pilots, in middle America, hitting trees or fields instead of apartment buildings.
On a side note, I don’t understand how that corridor can legally satisfy the requirements of an aircraft being 1000′ above and 2000′ horizontally from any structure. The corridor is less than 4000′ wide, so you are inherently within 2000′ horizontally from buildings on one side or the other.
Comments are closed.