Technique Geek: tailwind landings

More to it than meets the eye…

Both the FAA and NTSB tend to suddenly discover things that have long been a factor and make a big deal out of them. One or more accidents usually gets this ball in motion.

The latest hot button, from the NTSB, is what they choose to call tailwind landings. In what could have been a deadly serious accident, but wasn’t, an American Airlines 737 went off the end of the runway at Kingston, Jamaica. There were injuries but no fatalities. The 737 was pretty broken up about the matter.

You guessed it, the crew was landing with a tailwind.

American Airlines 737 off the runway in Jamaica.
A 14-knot tailwind led to a runway overrun for this 737.

The runway in use was 8,911 feet long. It was wet and there was a 14 knot tailwind component. The airplane touched down about 4,000 feet down the runway and the crew was unable to stop it in the remaining runway. It went through a fence, crossed a road, and stopped on the sand dunes and rocks just above the Caribbean Sea.

The regulations do not require any training on tailwind landings, nor does American Airlines provide any such training for its pilots. The airline does require that pilots perform an arrival landing distance assessment and that was not done by this crew.

I have studied other airline overrun accidents as well as similar general aviation accidents and there are always common threads. Too much speed at touchdown, whether from a tailwind or poor airspeed management, is almost always there along with wet surfaces and a touchdown well beyond the touchdown zone which starts 500 feet from the runway threshold.

As an aside, many (or even most) pilots do not fully understand the significance of runway markings. Those big thick white stripes that start 1,000 feet down the runway are aiming points, not touchdown zone markings. The touchdown zone markings start 500 feet from the threshold.

The American Airlines Flight Manual defines the desired touchdown point as within the first 800 to 1,500 feet beyond the landing threshold.

Clearly, if a tailwind landing is going to be attempted in any airplane, consideration has to be given to all factors.

When approaching my base at Hagerstown, Maryland, from the west I would request a straight-in to Runway 9 as long as the tailwind didn’t exceed 10 knots. The runway was uphill and the tailwind landing didn’t result in an abnormally long roll. The pilot’s operating handbook specified that it was okay to land with up to 10 knots of tailwind.

The airplane did feel different landing with a tailwind and I never managed smooth touchdowns though all were acceptable. I suspect that all airplanes are at least a little squirrely when landed with a tailwind.

The pilot’s operating handbook for your airplane should address the maximum acceptable tailwind for landing and give a method for calculating the amount of extra runway that will be required. On American 737s the maximum allowable tailwind is 16 knots. Southwest says 10 knots for the 737; five if the runway is contaminated.

There is one major factor in a tailwind landing that never seems to excite the curiosity of the investigators. It could have had a direct bearing on why this 737 crew wasn’t able to touch down until 4,000 feet of runway had passed beneath the airplane.

In theory the only increase in landing distance should be whatever is required to lose the additional groundspeed created by the tailwind. That would assume the same airspeed and altitude over the threshold and at touchdown as on an upwind landing. That, though, can be easier said than done.

The wind is almost always substantially stronger at, say, 2,000 feet than at the surface. At times it is quite a bit stronger.

For an example, say the wind at 2,000 feet is at 34 knots compared with the surface wind at 14, both acting as tailwind components for the approach and landing.

For the sake of argument let’s say the desired approach airspeed is 125 knots. At 2,000 feet the groundspeed would be 159 knots (125 plus 34 for the tailwind). If everything worked perfectly, the touchdown groundspeed would be 139 knots.

The twenty knot difference in the groundspeeds means the airplane would have to decelerate that much as it descends into the changing wind. This is easier done in a light airplane than in a heavier one, but it is something that a pilot has to think through ahead of time.

While decelerating on approach is something that had been done in heavy airplanes in USAF operations, it is probably not normally done in airline simulator training. There are even pilots out there who don’t actually believe this could be a factor.

A decreasing tailwind will, though, result in an increase in airspeed as the airplanes descends and if no accommodation is made for this, you might well land 4,000 feet down the runway. For this to happen you either have to be higher than the threshold crossing height or at an airspeed greater than Vref and with a contaminated runway either of those events should be cause for a go-around.

Back before we had GPS, a pilot had to use a little weather wisdom to anticipate this. There had to be some idea of the expected wind at 2,000 feet and the surface wind. Now there is nothing to it. Look at the groundspeed on the GPS while flying at 2,000 feet. If it is twenty knots greater than the airspeed, then you know how much you will have to decelerate while descending.

That changing wind with altitude is wind shear. Some pilots think of this only in connection with thunderstorms and there have been serious airline accidents caused by the failure of pilots to deal with storm-related shear. But a benign wind shear can also cause trouble and if a pilot doesn’t understand and anticipate this, he might find himself wondering what is going on with the speed on an approach. Wonderment is not one of the better sensations in a cockpit.

Tags from the story


  • Mastering a tailwind landing should be practised. It can be a very useful option on approach. Straight-in ILS allows lower minimum than circle-to-land: Just request a straight-in instead of going missed.

  • I fly a 737-800 just like this one and can tell you they land fast (to avoid tail strikes) and don’t have very good braking ability on wet, ungrooved runways such as the one in Kingston this day. Throw in a pilot trying to make a decent touchdown and floating it a little too far, and the recipe for disaster is set.

  • The Southwest Midway crash was an example of both pilot and controller stupidity.

    I don’t know the facts in this case, but I suspect they won’t be much different.

  • Landing with a tailwind should always be a last resort. Even if the POH gives you a Max tailwind limit, that is just telling you what the aircraft has been certified to do. There are many techniques one can use to assist if you have to make a tailwind landing; including reference ground speed. Pilots should ask for a runway change or an opposite direction approach; it may mean more time in the air, but it probably will save you. With over 5000 hours in a variety of aircraft, i have never found a tailwind approach to be a preferred option. Sometimes you have to do it for wx reasons as mentioned, but you should never do it for convenience….yours or the controllers.

  • Is it me or it’s sound like another basic skill / knowledges pilot skipped? i’m sure there is a few procedure to re-configure the 737 for a go-around 3500′ after touchdown zone, but still?! back here there is snow, we train private and commercial pilot with the rule: at 1/3 of the runway is there a doubt you don’t have at least 1 + a half of your groud roll left, you go around. with 4000′ of runway left, I guess it can be called a short-field landing for a 737-800?

  • The statement “The twenty knot difference in the groundspeeds means the airplane would have to decelerate that much as it descends into the changing wind.” seems off. I don’t think the A/C knows it had a tailwind and the pilot does not have to reduce airspeed on descent. The additional ground speed just has to be lost on the ground.

    • There have been countless accidents caused by wind shear and that is what happens when the wind changes as you descend. A pilot who doesn’t understand this will wonder what is going on with the airspeed and vertical speed during a descent in changing wind.

      • Accidents caused by wind shear are often disastrous, deadly events, “countless” maybe a little overstated. The term “wind shear” CAN be defined as ANY change in wind speed and direction but it’s more often used in relation to wind changes associated with weather features such as thunderstorms, squall lines, strong cold fronts, hurricanes, etc. and sure clear air turbulence. While all pilots MUST be aware of and understand sever wind shear, micro burst, etc., how to react to them and hopefully how to avoid them, they should not think that they occur regularly in decent weather when recreational pilots tend to choose to fly. Normally there will be some warning signs; lightning in a thundercloud, hail, generally strong, variable low level and surface winds, weather features often associated with good days to stay in the hanger, quite unlike this mourning. Let’s go flying! The weather is perfect here in NW Florida. Probably have a few wind changes but almost no chance of a dangerous wind shear. Yesterday was a different story. Sorry for the hair splitting.

        • Thad,

          Unfortunately you have hard time understanding the wind component difference and altitude.
          Very simple meteorology basic related to Coriolis force. Wind shear was explained very clearly regarding descend by R.C.

  • “A decreasing tailwind will, though, result in an increase in airspeed as the airplanes descends…” is how albatrosses stay aloft without flapping their wings for days on end. It’s known as dynamic soaring. Albatrosses convert that increased airspeed to altitude in order to make another glide back down through the wind gradient, repeating this as often as they’d like. Sailplanes don’t have the turn radius to take advantage of it and hang gliders don’t have the glider performance to use it, though both sets are earnestly trying to capitalize on it. For the crew of this airliner, however, I suppose it was the last thing they wanted.

  • An abrupt horizontal wind sheer(a rare occurrence, I think) may do some odd things but generally changes in winds change ground speed and direction and the albatross is probably more concerned with up/down drafts and waves over hills, etc. as they affect its altitude.

  • All commercial aircraft have a talewind limit, whether it be aircraft limitation of company operations limitations. Each runway and each condition is different and when within “legal” limits, the captain (or PIC) has to make a judgement call. He has to consider his personal limits, the weather at the time, the aircraft, runway condition and, of course, alternate availability if you choose not to land. (there is probably more to consider too). That is what the Captain is paid for.

  • Speaking strictly WRT smaller GA, I have found that limited tailwind components are easily handled so long as the pilot remains aware of all the basics: air, vertical, and ground speeds and the changing relationship of aircraft to the target landing zone. Remember that the circumstances of that landing will shape your decision to make a tailwind landing (like the time my girlfriend HAD to land RIGHT NOW, and the tailwind landing put us right at the “correct” end of the runway…) As PIC we’re expected, and required, to make the smart decision all the time.

Leave a Reply

Your email address will not be published. Required fields are marked *