An Investigation of PG Landings in Wind Shadows

An Investigation of PG Landings in Wind Shadows


by Dan Harrison


Have you approached a landing with ground streamers showing a light, opposing headwind only to land in a tailwind or in sink? Did you chalk it up to an LZ thermal or just to fickle winds? This article investigates some potentially hazardous landing conditions in winds too light to create fully formed rotors. We’ll see how a pilot may be fooled by a backwash on the ground induced by a terrain feature underlying the true wind-stream.


Unexpected Conditions

Let’s say you’re approaching a Landing Zone (LZ) and streamers show light winds. You’re expecting an easy landing into a light, opposing breeze. Yet, as you prepare to land, your speed increases and you suddenly realize you’re in a tailwind. If you‘re inexperienced you may make the instinctual mistake of braking to slow to what you expected your landing speed should be in a light headwind.

Normally, this isn’t too dangerous. You descend a little faster and have a harder landing. But if you’re turning onto final as you encounter this tailwind you may inadvertently stall your wing on the inside of the turn while having swung in a pendulum motion to the outside of the turn. Then you are swung back and down slamming into the ground. Let’s review what’s happened. 

You turn into a light headwind shown by a ground streamer or other indicators, but it feels like you’ve entered a partial vacuum. In fact your wing is in a tailwind at the altitude of your glider about 25 feet or so above you. This tailwind is much stronger than the backwash on the ground and flows in the opposite direction. By turning in line with this tailwind, your ground speed increases, but your airspeed relative to this tailwind drops dramatically. 

The wing on the inside of the turn stalls. Having swung toward the outside of the turn, as your inside wing stalls you now pendulum back down toward the center, and as the glider drops like a rock, you’re slammed to the ground in an awkward position completely unprepared for a Parachute Landing Fall (PLF). Depending upon the specific terrain, wind conditions, and your choice of maneuvers, this effect can be pronounced and very dangerous. 

I observed what might be attributed to these conditions when a new P2 pilot at Ed Levin Park in Milpitas, California, broke his leg with a compound fracture. Previously I had observed a similar accident with a more experienced P3 pilot turning final for the (old) main LZ at Potato Hill in Fouts Springs, California. This accident resulted in a broken pelvis and clavicle. 

Is there something we can learn from this besides the well-known mantra: keep your speed up and don’t turn close to the ground? Both of the accidents described above involved low turns on approach. But in both cases the pilots appeared to be reacting to conditions that were not as anticipated. The pilots appeared to be trying to escape these conditions—too late and incorrectly—without any concept of what was really going on. 

I fly cross-country and occasionally have to land in tight spots. Worse yet, I sometimes land in LZs that just don’t recognize this mantra. There are situations (for example, LZ thermals) where if I kept my speed high and didn’t turn I’d have overrun the LZ—into a barbed wire fence or trees or worse. Let’s look at what’s going on so that we can learn to modify our landing approach to reduce the risk of injury to ourselves, to spectators and to our LZ sponsor’s property. 


The Pre-Rotor Terrain Induced Backwash 

Any obstruction to airflow will produce a Wind Shadow. What is not commonly known is that within the lee of this obstruction (within the Wind Shadow) a low pressure sets up that will pull air from all directions. As we’ll see, with the obstruction in front, air is pulled from the sides and from behind. I’ll call this secondary airflow, which is generally opposite to the dominant wind-stream above, a Terrain Induced Backwash (TIB). As we’ll see, a TIB represents an early stage of rotor development before a rotor is fully formed. (You can see this illustrated in the You Tube videos in the first Technical Insert Box below.) 

To see how we can learn to recognize and mitigate the risks of landing in these situations, let’s look at the conditions that can induce a TIB and that may lure us into a dangerous landing approach. Then we’ll investigate a modified approach that will allow us to avoid the more severe consequences of a TIB.

 TIB on a Wing

Conditions Favoring a TIB

As pilots we learn that the airflow above a wing in flight flows faster than the airflow below, resulting in a lower net static pressure above the wing. This produces lift. Even a blunt object can generate a pressure differential in this way, albeit with enormous drag. 

Suppose we as observers are on a hill on the South side of a broad flat LZ (refer to the photo) and looking North over the landing field. Several streamers (represented by the red arrows in the photo and pointing to the left) are set up just above the ground to indicate wind direction. 

A wind-stream blows across the LZ from our left well above the ground and opposite to the direction indicated by the streamers. This wind-stream is invisible to us as it is too high or too light to produce noticeable movement in the pine trees. Or it produces light movement, but we cannot tell if it is from the right or from the left, so we rely on the streamers in the field, which are pointing toward the West, to incorrectly determine its direction. 

Consider first how this wind would behave in a barren, flat terrain. As the wind approaches the ground, it slows due to ground resistance. This reduction in velocity increases static pressure forcing the wind-steam back up until the wind-stream reaches an altitude above the ground at which there is no vertical pressure differential. This is why a cool surface wind will remain a few feet above the ground and follow the contour of the surface. 

In barren, flat terrain we expect streamers and other ground clues (blown dust, debris or grass) to reflect the true wind direction and reduced wind speed close to the ground. But when the terrain around an LZ is not barren and flat, as in the photo above, these ground level steamers may mislead us even when the conditions do not appear to pose a risk of rotor development. Let’s consider what may happen in the terrain shown in the photo above. 

In this terrain we have a West wind (invisible to us) crossing raised rocky ground with scattered pine trees just before dropping to cross a road into a flat valley clearing. (Whenever a wind-stream crosses an obstruction a pilot in the lee of this obstruction is said to be in a Wind Shadow.) 

If this West wind-stream had a higher velocity, we would observe significant movement of the pine trees and would anticipate that a rotor may be setting up just beyond the trees and road. A rotor would explain the opposite direction of the streamers, and we would know to land further to the East in the valley clearing to avoid this rotor. 

But in this situation the wind-stream is too light or too high to produce significant movement in the pine trees. So we might assume that the wind is from the East, as the streamers appear to indicate, and may actually be setting up a rotor on the far Eastern side of the valley to be avoided at all cost. 

Also in this LZ, pilots landing in the valley are going to be picked up at the road on the left. This encourages pilots to land close to the road whenever it appears safe. A pilot approaching the valley from the left (from the West) with ground streamers pointing toward the left (implying a light East wind) may be induced to land close to the road just over the pine trees. As the terrain is dropping, this may not appear to be dangerous. Unfortunately, the situation is deceptive. 

The pilot will descend through a tailwind on approach that will significantly increase his ground speed. But as his glider enters this tailwind, his velocity relative to the tailwind will drop rapidly just as his ground speed increases rapidly. He will lose lift and drop rapidly toward the terrain where he is committed to a landing straight ahead. If he has not given himself enough of a safety margin, a wingtip may snag a pine tree just short of the final drop into the valley. In the worst case this could drop him into traffic on the road.

The pilot may make two mistakes in this situation. First, as his ground speed increases in this tailwind, he may, if inexperienced, use brakes to slow his ground speed. But his velocity relative to this above-ground tailwind dropped as soon as he entered the tailwind. Reducing his ground speed using brakes after entering the tailwind will further reduce his velocity relative to the tailwind and may result in a low altitude stall.

The pilot may make a more serious mistake if he attempts a low altitude turn crosswind too soon. The probability of a stall of the inside wing in this situation is very high. This is what I postulate caused, or contributed to, the two accidents described at the beginning of this article. A cross-wind turn can be made safely only after your glider’s speed has been allowed to increase relative to the tailwind. You may be on the ground before this happens.

In this situation commonly called a Wind Shadow, we should always aim for a landing at least a third of the way into the LZ even when we appear to be landing against a light headwind. The TIB theory provides a strong foundation for this rule and counters the excuse for landing close to the road where you can save yourself from a slightly longer walk out to the pickup area. 

 Yosemite LZ

The Aerodynamics of a TIB and Rotor

So how is it possible to be deceived in this way? Why do the ground streamers show a light breeze from the East under a West wind-stream when this wind-stream is too light to generate a rotor in this terrain? How is this backwash created without the formation of a rotor?

In fact, every airfoil will generate this backwash just before the velocity reaches a level high enough to create a full rotor. For an airplane in flight, the air for this backwash comes from behind and below the wing (rather than from the airstream above the wing as in a full rotor) and can easily be seen in aero-tufting experiments on aircraft in flight or in wind tunnel experiments using smoke. (This can be seen in the You Tube videos referenced in the 1st Technical Insert).

In the photo above the surface wind- stream drops into the LZ from the left after crossing slightly higher rocky ground with shrubs and small pine trees. The lower portion of this wind-stream is forced to increase its speed by the air mass behind in order to cover the greater distance caused by this contour. As the velocity—and dynamic pressure—increase, the static pressure drops (as it must to observe the law of Conservation of Energy for an adiabatic process—2nd Technical Insert).

This drop in static pressure sucks in air, below the wind-stream, from the sides and from the downwind direction (which is on our right or from the East). As this backwash increases we see streamers near the ground pointing toward our left, the direction from which the above ground wind-stream is actually coming. 

The pilot, seeing the streamers pointing West without any indication of rotor, may set up to land toward the East in what is thought to be a light opposing headwind. 

He will then have to descend through a tailwind much stronger than the opposing light backwash identified by the streamers on the ground. And if the pilot approaches final from the North or South, he will be turning into a tailwind, or descending through the tailwind immediately after his turn as his ground speed unexpectedly increases rapidly at low altitude on final. This will be alarming, and may create confusion.

To reiterate why this backwash occurs, recall that our wind-stream (cooler and hence, more dense and heavier, than the surrounding ambient air) dropped simply because it passed over a terrain feature that allowed it to drop. This feature could just as well be a small rise in the terrain, a farmhouse or a tree line. 

This feature acts as an obstruction preventing air from flowing from the upwind direction. And if the obstruction and overlying  wind-stream have any lateral breath, air flowing from the sides will not reach the center of the stream. Hence, the majority of this air must come from the downwind direction, thus creating a backwash. 

This backwash is directed toward the “obstruction”, and remains below the much stronger wind-stream that is flowing from the opposite direction above this backwash. Note that the backwash is not created by ground friction, but by the restriction of airflow feeding the low-pressure area that has been established on the lee side of an “obstruction”. Hence, this LZ is said to be in a West Wind-Shadow. 

Thus far I’ve been careful to define the conditions on the ground as sufficiently mild so as not to suggest the possibility of a full rotor formation behind this low obstruction. As the wind-stream velocity—or height of the obstruction—increases, the possibility of a rotor may be indicated by strong wind in the trees, or in the absence of trees, by windblown dust or other debris. Still, detection of these windblown indications may come too late, and a pilot, fooled by ground level streamers in the backwash, may find himself committed to attempt a landing in a fully formed rotor.

 Technical Description

Landing in a TIB

So how do we read these clues to recognize the conditions and improve the safety of our landing approach? Air at ground level will be drawn towards the obstruction. This piece of information is critical to developing a safer landing approach, as you may have no other indication of the true wind-steam direction at your wing’s altitude 25 feet or so above the ground. It also warrants a brief observation about the generation of thermals on the upwind side of a tree-lined LZ or other open area, our first corollary to the TIB Theory.

If you have an LZ sprinkled with streamers about an “obstruction”, a TIB will cause these streamers to point toward the obstruction from lateral positions as well as from directly “behind” or downwind of the obstruction. The presence of the obstruction is a strong clue to the potential development of a TIB. 

But it also looks very much like the streamers are pointing toward a thermal source initiated by this obstruction. And, in fact, that may be the case. If conditions are right, this backwash can use the obstruction to build enough force to break through the overhead wind-stream (if the wind-stream is not too strong). This is how a tree line, or other low obstruction, can trigger a thermal on the upwind side of an LZ! 

So, how do you land safely in these situations? You may not know that there is a strong wind-stream above the ground, but you do know that the winds on the ground are very light because you’ve identified some clue such as a streamer, movement of vegetation or dust. And you know its general direction as well from these same clues. Finally, you (usually) will know if these ground clues are pointing toward an “obstruction”. So, how do we set up our landing approach?

We don’t always know if we are dealing with an LZ thermal, a TIB or just a light headwind. So whatever approach we take needs to work for all three of these situations. Consider what will happen if you choose to land with a substantial crosswind component in each of these cases.

Say you approach the landing with a 30 to 45 degree offset from what appears to be a light, opposing headwind as indicated by streamers or other clues on the ground. (The key word here is light. In a strong or even moderate headwind, you must set up to land directly into the wind. Even a slight offset in a high headwind landing can be dangerous. In a strong headwind you will need to land in control so that you can immediately “kill” and control the glider so you will not be dragged across the terrain.) 

Case 1. You have turned crosswind to what appears to be a light headwind. If you turn out to be in a thermal you will be lifted, and with sufficient altitude, can immediately turn slowly back towards it after evaluating two risks that this situation presents. 

First, there is a risk of getting dumped out of the thermal at a low altitude. If you are in a turn close to the ground when this happens, you can stall your inside wing and pendulum into the ground. The second risk is overrunning the LZ. But your initial crosswind approach has already positioned you to mitigate both of these risks.

When you are close to the ground and find yourself in an LZ thermal the best option is to land straight ahead if you will not run out of landing area, crosswind at this point. But if you’re going to run into an obstacle and must change direction, then turn slowly into the wind. Turning into the wind will reduce your chances of getting dumped out of the thermal. In addition with a turn into the wind, the wing on the inside of the turn will still have lift as the outside wing drops out of the thermal. This will greatly reduce any resulting pendular motion. Note that your initial crosswind approach will have given you additional maneuvering room to deal with a short LZ. If instead of landing you continue to climb, you’ll continue flying or set up for another landing when the thermal abates.

Case 2. If this ground-level headwind turns out to be a TIB, by approaching the landing crosswind you will have significantly reduced your exposure to the tailwind above, and you can land straight ahead with a significantly lower decent rate and ground speed than you would have otherwise experienced. 

Case 3. If the headwind turns out to be nothing more than a light ground wind, you will land straight ahead in a light crosswind.  

In summary, if you do not know what the winds at wing level above the ground are doing, but you do know that the wind on the ground is light in terrain that might produce rotor in more severe conditions, you may be landing in the TIB of a Wind Shadow. If in fact there is no tailwind, at most your landing speed increases only very slightly by landing at a 30 to 45 degree crosswind angle to the headwind. Just be prepared to land in a light crosswind as opposed to a strong tailwind. As the headwind is light a Parachute Landing Fall (PLF) will probably not be necessary. But just in case be prepared for a PLF into the headwind.

If you do encounter an unexpected tailwind, your crosswind approach will have positioned you much better to gently turn even more crosswind (with sufficient altitude), or simply to land straight ahead with a significantly lower ground speed and lower descent rate than you would have otherwise experienced. But this may still be a high-speed landing with a higher than normal descent rate. So how do you land in a tailwind? 

Jack Guthrie in An Inconvenient Truth Regarding PLF’s,, 2010-03-02,, following an analysis of the physics of a parachute landing, suggests that a pilot should “assume the PLF position, fly the parachute and slide on one hip in the event of a forward motion, high speed landing”. The pilot should, “try to absorb as much as possible of the downward impact with their feet but lean back in the harness. Under no circumstances should the person allow themselves to be thrown head first.”

Note that a PLF requires “falling” to the right or to the left, not straight ahead. So the 30 to 45 degree offset you have already set up will allow you to execute a PLF (falling forward), or Jack Guthrie’s maneuver (sliding back on your hip), without any further change of direction.


A TIB Corollary and Application

A characteristic of any good theory is an ability to draw generalizations that expand upon the theory or provide for applications. We have already seen how a thermal can be initiated at a tree line, or other obstruction, on the upwind side of an LZ. This thermal results from a strong TIB breaking through the prevailing overhead wind-stream and is fed by heated air on the ground—our first corollary to the TIB theory. 

There is another very subtle, but potentially common, situation that I’ll postulate can create a Terrain Induced Backwash (TIB) without a full rotor setting up. On flat open ground in high desert terrain, for example, a light to moderate cold downdraft may strike the ground and bounce, then level off just above the ground. 

The position at which the downdraft bounces creates an effective obstruction (due to the mass of the wind-stream itself). The contour of the bounce may result in a backwash under the bounce directed back toward the “obstruction”, which is the point at which the descending wind-stream strikes the ground. 

So it is possible to have this situation set up in a broad, flat, open LZ with no visible terrain features. All we have in this situation are clues on the ground—whether due to streamers or light movement of leaves or dust kicked up by other pilots. (Stronger downdrafts called Fallouts, Downbursts or Microbursts described by Dennis Pagen in his book, Understanding the Sky, are likely to overwhelm this effect. But more moderate downdrafts can be expected to be far more common.)

Now consider another example that results from a Wind Shadow. As part of a site intro for the Inspiration Point — or Inspo — West-facing launch in Orem, Utah, I was briefed on a peculiar situation at a bailout LZ. (This is a cross-country P4 rated site.) With little lift that day I chose to land at this bailout, a local school. I had been briefed that this LZ was notorious for consistently dropping the pilot from a few feet above the ground just before touching down. This single story school presents a low obstruction to a typical afternoon wind from Utah Lake to its West. There are no streamers, but the pilot has been approaching in a headwind, so assumes he will land in a headwind if he lands against the West wind. 

On final, just before touching down, I dropped rapidly and my speed increased making for a harder, faster landing than anticipated, but consistent with the pre-flight briefing I’d just received. Did I land in a backwash created in a wind shadow by the West wind flowing over the schoolhouse? 

This bailout is surrounded on all sides as follows: a schoolhouse to the West (the direction from which the wind is coming), tree lines to the North and South, and a tree line with power lines to the East. As the wind is generally from the West, turning one-eighties (180’s) may be required just inside the tree line and power lines to the East of the LZ just before lining up on final against the West wind. 

So what happens if on approach we assume a TIB (backwash) will have set up, and we decide to land crosswind? Conversations with local pilots reveal that experienced pilots have often landed in this small LZ crosswind to the prevailing West wind with the same increase in descent rate close to the ground and with increasing ground speed. What could cause this increase in descent rate and ground speed in any direction? Is this a TIB or something else?

The small size of the LZ, with obstructions on all sides, means that the low pressure set up by the Wind Shadow will have difficulty pulling air from any direction. There may not be a prominent direction from which air can be pulled to feed the low pressure, and hence, no clear backwash. But the low pressure, because it’s difficult to fill, will be unusually strong relative to the wind strength, and will cause your wing to drop quickly. And this increased descent-rate will translate into increased airspeed toward whatever direction we are landing.  

In Part II of this series, An Investigation of PG Landings in Strong Winds, we’ll begin to explore landings in conditions that produce rotors, strong tailwinds and other strong winds and turbulence. As weather conditions may change between the time we launch and land, and as we sometimes land not where we intended, we can expect to be forced to deal with a high wind landing at some point. Hence, this article will be about landing in more dangerous situations.