Please wait...


Birds and dynamic soaring

Shearwater2Can a bird use dynamic soaring in a steady homogenous wind?

Philip L Richardson
Woods Hole Oceanographic Institution

Colin Taylor's website and his BOU blog article 'Dynamic Soaring' both argue that energy-neutral soaring flight can be achieved by a bird flying in a steady homogenous horizontal wind.

To my mind dynamic soaring is a technique in which birds exploit the vertical gradient of wind velocity (wind shear) and also temporal variations of wind velocity for sustained energy-neutral soaring (see references below). In this technique a bird climbs up across the wind shear layer headed into the wind, turns and descends down across the wind shear layer headed downwind, and then turns to head upwind again. Albatrosses are known to use dynamic soaring to fly very long distances without flapping their wings.

There are no documented studies of albatrosses using a steady homogenous horizontal wind for sustained soaring, which raises questions about how energy could be gained in Taylor's Windward Turn Theory. In order to try to understand this, I carefully read material in the website and discussed it with Colin Taylor.

In the following, I summarize problems with the Windward Turn Theory that invalidate its conclusion about energy-neutral flight.

The main point is that if one assumes a steady homogenous horizontal wind then sustained soaring cannot be achieved. This is because documented mechanisms for energy-neutral soaring are excluded. Specifically, a homogenous wind excludes gradients of the wind velocity (wind shear), a steady wind excludes temporal variations (wind gusts), and a horizontal wind excludes updrafts. Thus, dynamic soaring is excluded as is soaring using updrafts. Therefore, the answer to the question posed in the title must be no.

The shortcomings of the Theory are briefly discussed below. In this discussion drag is assumed to be negligible. This simplifies the physics since drag always reduces energy, and by assuming that drag is zero one can calculate energy gained from flight maneuvers. A second assumption is that total energy, consisting of kinetic energy (KE) plus potential energy (PE), remains constant for a bird in flight. A third assumption is that in the following the term 'wind- means a steady homogenous horizontal movement of the air.

There are two main flaws in the Windward Turn Theory. The first is that it uses the velocity of the bird relative to the ground (ground velocity) in the calculation of KE, which is then incorrectly assumed to balance PE. A bird's ground velocity (or ground speed) is the sum of the bird's velocity through the air (airspeed) and the wind velocity. Airspeed is the usual and correct term in calculating KE in the balance with PE in soaring flight. This can be seen in the example of a bird accelerating as it glides downward. The bird's airspeed increases accelerated by gravity, and height decreases. As sailplane pilots describe it, height is traded for airspeed. The trade is not linear since airspeed is squared to calculate kinetic energy, and drag must be ignored. The wind has nothing to do with the balance between airspeed KE and PE. There are many examples indicating why this is true.

Consider a bird flying in a level circle at constant airspeed in a wind. The bird's ground speed increases when the bird turns and heads downwind. The increase in ground speed is due to the bird's acceleration in a downwind direction caused by the lift force on the bird's banked wings. The KE of the bird is increased by the work done by the wind on the bird (force times distance). In this case, the work is done by the downwind component of lift times the distance downwind that the bird is advected by the wind. As the bird turns upwind its ground speed and KE decrease due to the work done by the bird on the wind. The point is that the bird cannot gain or lose height in the turn as long as its airspeed remains constant despite large changes in ground speed KE.

A simpler formulation of the equations of motion would use a reference frame moving with the wind velocity. This eliminates wind from the solution and avoids the work of the wind on the bird, etc, as when using a reference frame attached to the ground. In this simpler case the balance between airspeed KE and PE is obvious.

Richardson Albatross and UAV

The Windward Turn Theory incorrectly assumes that there is a height loss (loss of PE) when the bird turns downwind and its ground speed and ground speed KE increase. And the Theory incorrectly assumes that there is a height gain when the bird turns upwind and ground speed and ground speed KE decrease, despite the bird's airspeed remaining constant in each turn. The use of ground speed in the calculation of total energy results in huge KE changes and corresponding unrealistically large apparent height changes, much larger than when airspeed KE balances PE. The Theory claims that there is a large increase in height as a bird turns from headed fast downwind to headed slowly upwindÔÇöthe windward turn of the Windward Turn Theory. However, this height is lost in the following leeward turn.

So, how could energy increase in the Theory since it excludes documented sources of energy? I believe that the Theory accidentally omits important terms in the energy equation or incorrectly calculates their values. As a test of this, I developed a model that simulates the Theory's flight pattern of two linked 180 degree turns in a mean crosswind direction and incorporates the conservation of total energy using ground speed to calculate KE, which then balances PE (as in the Theory). My calculations found no net gain of energy in the Theory's flight maneuver. I conclude that there is no net gain of energy and that the Theory cannot be used to explain energy-neutral soaring flight.

In summary, two mistakes resulted in an erroneous apparent energy gain in the Windward Turn Theory. The first is the use of ground speed to calculate KE, which is then incorrectly used to balance PE. The second is the incorrect calculation of energy terms in the Theory, which leads to an apparent net energy increase in the flight circuit. Without these mistakes, there is no net increase of energy in the Theory's flight maneuver. Energy-neutral soaring flight is not possible in a steady homogenous horizontal wind since dynamic soaring is excluded as is soaring using updrafts.


Deittert, M., A. Richards, C. A. Toomer, A. Pipe, 2009: Engineless UAV propulsion by dynamic soaring, Journal of Guidance, Control, and Dynamics, 32 (5), 1446-1457.
Lissaman, P., 2005. Wind energy extraction by birds and flight vehicles. American Institute of Aeronautics and Astronautics Paper 2005-241, January 2005.
Rayleigh, J. W. S., 1883. The soaring of birds. Nature 27, 534-535.
Richardson, P. L., 2011: How do albatrosses fly around the world without flapping their wings? Progress in Oceanography 88, 46-58.
Sachs, G., 2005. Minimum shear wind strength required for dynamic soaring of albatrosses. Ibis 147, 1-10.

Philip Richardson is Scientist Emeritus at the Woods Hole Oceanographic Institution where he has studied ocean currents for many years. He has enjoyed a lifetime interest in flying (and sailing) and has piloted sailplanes and powered aircraft. His interest in dynamic soaring began when he observed albatrosses in the Southern Ocean soaring upwind for long times without flapping their wings. He developed a fairly simple model of dynamic soaring, which simulated albatross soaring including upwind flight (see references). The model was also used to simulate the fast flight (up to 500 mph) of radio-controlled gliders dynamic soaring in winds blowing over mountain ridges. Recent work promotes the concept of a fast robotic albatross that can soar over the ocean waves like an albatross but much faster than a real albatross because of stronger wings. Results of these two projects were published in R/C Soaring Digest in 2012. Email Richard.

Blog posts express the views of the individual author(s) and not those of the BOU.

Images: Cory’s Shearwater © Steve Dudley; Composite image of Black-browed Albatross and Kinetic 100 radio-controlled glider © Philip Rciahrdson


Blog with #theBOUblog

If you want to write about your research in #theBOUblog, then please see here.

29 comments on “Birds and dynamic soaring

  1. Ross Whittle says:

    I sympathise, Richard.

    Several people, including me, have corresponded with Colin Taylor attempting to point out the glaring errors in the physics of his theory, but to no avail.

  2. Peter Hirst says:

    I don’t know about his physics, but the phenomenon is real.

  3. Peter Hirst says:

    I should add that I believe Albatrosses derive the majority of their free energy for sustaining continuous gliding flight by flying in ground effect very close to heaving swells. However, I am also sure that they are able to derive some energy whilst turning and climbing from crosswind towards the eye of the wind – And that wind gradient is not a factor in that derivation. The action of turning within the quadrant I have specified derives energy even from a steady wind.

  4. Ross Whittle says:

    Ground effect reduces drag and allows longer glides, but it does not allow indefinite gliding distance.

    No, turning from a cross to a head wind does not increase the ability to climb- if you are heading crossing, you are already drifting down-wind at the speed of the wind- you do not gain any extra airspeed by turning into wind.

  5. Peter Hirst says:

    I didn’t say ground effect provides for indefinite gliding – I am saying that an Albatross is able to employ ground effect to great advantage in combination with slope lift produced by the wind over heaving swells and waves. And that I believe it derives the large majority of the energy it requires for continuous gliding in this way. Also that it is this energy that provides for the pull up.

  6. Peter Hirst says:

    Nor did I say that additional airspeed can be gained by turning into wind. However, I am saying that I believe energy can be derived in a turn towards the eye of the wind that can preserve airspeed when otherwise it might be decaying. In an aircraft you need to be close to the stall and have a wind that is equal to perhaps as much as half your stalling speed to feel it.

  7. Ross Whittle says:

    I’m not sure what it is you are suggesting Peter, but, no, there is no way to derive energy from a turn into wind.

    Airspeed decays as a function of drag, which is proportional TO airspeed- that is, speed relative to the air.

    What that air is doing relative to the surface has no bearing to the relationship between a bird or aircraft AND the air.

    If there is such an effect (which would defy Conservation of Energy) how close you are to stall, or what relationship the wind speed had to stall speed would be irrelevant.

  8. Ross Whittle says:

    As to your first response, yes slope lift over swells might well provide some part of the lift the bird utilises.

    However, there is no doubt that wind gradient CAN be used as a source of energy- whether it is what an Albatross utilises is for those expert in the subject to debate.

    Mr Taylors original article and web-site uses VERY erroneous physics to suggest gaining energy from wind gradient is not possible- in this he is absolutely mistaken.

  9. Peter Hirst says:

    Ross I suspect you maybe a pilot – Anyway I have been aware for nearly forty years that what I am describing is a factor that is not acknowledged in aviation.

    Nevertheless it is by belief on account of substantial personal experience that drawing/turning an inclined foil from crosswind towards the eye of a steady wind – which by inference necessitates an acceleration of greater than 1G – temporarily creates the effect of an increasing local gradient over the foil. Continue through the eye of the wind and a decreasing local gradient will be experienced – In normal aircraft operations it is a negligible and barely noticeable factor. Fly low and slow and start turning in a strong wind and it is very noticeable to me. It feels safe manoeuvreing towards the wind and dangerous coming off of it. And I’m not talking about the appearance of the ground and the aircraft’s speed over it – I’m talking about reassuring and/or worrying feedback through the controls in respect to stall speed. As to the reason – a turn naturally comprises an acceleration. My thinking is that accelerated and decelerated flight conditions are effected in a steady air-stream in a way that unaccelerated flight clearly is not.

    Regarding wind gradients – of course energy can be derived from them, but I don’t believe there is anywhere near sufficient wind gradient between the surface of the ocean and fifty feet for birds to exploit it as their main source of energy for continuous gliding flight. I used to windsurf a lot, but I have never felt a wind gradient between the top of my sail and the bottom of it – indeed tell me who the sail manufacturer is who is making sails for any sailing craft of any size and taking into account potential wind gradients between the top and bottom of their sails? …Meanwhile on the subject of sailing it is also my belief that the phenomena I have described in the first paragraph assists sailing craft in tacking.

    As for experts and the means by which Albatrosses and Petrels derive energy for free flight – I THINK THE ONLY EXPERTS ARE THOSE BIRDS.

  10. Ross Whittle says:

    Hi Peter,

    I’m sorry, but what you suggest simply defies very basic newtonian physics. What you have to realise is that your speed relative to the ground is equaly effected by wind what ever direction you are going.

    If you are pointing exactly crosswind in, say a 15 knot westerly, you are ALREADY drifting downwind at EXACTLY 15 knots.

    An airfoil simply responds to the air around it- whether that air is moving relative to the ground makes no difference.

    Think of it this way- imagine you are on a train travelling along the tracks, and in front of you is a fish tank. The water in the tank is travelling at the same speed as the train.

    Will the fish in the tank feel a surge of water over i’s gills if it turns from a path travelling across the direction of travel to one in the same direction? It is turning from a cross-current to the “eye” of the current- will it feel this? Is it energy that could be used?

    Of course it is not. Now imagine an enormous fish tank full of air, with an Albatross inside it- this is effectively what a bird in flight in a wind IS.

    There are perceptual illusions that can make it SEEM like there is an effect- look up “Apparent slip”- I have a feeling this is the phenomena you are describing.

    As to your doubts as to the potential energy gains from wind gradients- have a look at the accidents it has caused it Aircraft, or research the fact that the slight difference in free-stream between the wind over a hill and that in it’s lee have allowed model gliders to achieve speeds of over 400MPH!!

  11. Ross Whittle says:

    AS to sailing, the big difference is that you are using TWO media- water and air- and the interaction between them complicates them a lot. Imagine if your sailboard had no keel and no friction on the water- you couldn’t tack, you couldn’t hold a reach- ALL you could do is move with the wind at the winds speed.

    As to gradient, I’ve done some ballooning. As you know, a balloon moves at exactly the speed of the wind around it- yet when flying low I’ve felt a breeze on my face- Why? because there is sufficient gradient between the top and bottom of the balloon for the envelope to be operating in a different wind than the basket.

    Surface boundary layer is well understood and quantified- and certainly sufficient to allow an Albatross to remain airborne.

  12. Ross Whittle says:

    Here is EXACTLEY an article explaining the relevance of wind gradient on sails-


  13. Peter Hirst says:

    Hi Ross

    The train carriage thing is aviation folklore that is not suitable for discussing acceleration and deceleration in flight. The theoretical Albatross in your train carriage would be flying around in technically still air even when the train is doing 40 mph. So for instance it could still takeoff or land in your train carriage in any direction it pleases – as it would normally in still air in the outside World.

  14. Peter Hirst says:

    I view sailing as vertically oriented flight at the interface of two different fluid mediums. In the case of my windsurfer – it has two wings. Comprising a sail that flies in the air and a fin that flies in the water. The board seals the interface and provides something to stand on. Sailing is a form of flying.

    I salute you Ross for finding someone who is addressing the relevance of wind gradient up and down his sails. And I am looking at that, but in the meantime would like to point out that he is representing his data for model yachts and the wind gradient solely for 0 – 10 knots of wind.

  15. Ross Whittle says:

    Hi Pete,

    Yes, the train carriage would represent still air.

    So too does a bird in the wind once airborne.

    Once it is no longer in contact with the ground, it’s speed relative to the ground plays no part in the dynamics of it’s flight. It’s accelerations are equal in any frame of reference, so suggesting it’s changing ground speed is somehow more relevant than it’s stable airspeed is doing what is known as “Invoking an absolute frame of reference”, which is a common misunderstanding of the basic physical principle that velocity is a relative value.

  16. Ross Whittle says:

    The huge difference between sailing and flying is that there are two mediums involved.

    You cannot, for example, use a sail to power a balloon, as there is no water to anchor the craft and allow any relative airflow. A balloon simply drifts in the direction of the wind at the speed of the wind, and can only be controlled in altitude.

    Similarly, an aircraft fly’s IN the air- and it’s path over the ground is a simple vector addition of it’s path through the air and that airs path over the ground (wind!!).

    As such, the train analogy is not folklore- it is a valid analogy- can you explain who a birds interaction with the air is influenced by it’s movement relative to the ground?

  17. Peter Hirst says:

    Ross I had thousands of hours flying for a living at a time when a former Hawker test pilot and UK 15 metre national Gliding champion who I was often working with got me interested in flying radio controlled model gliders. After crashing lots of them I finally became quite proficient with slope soaring models – By the way I have never crashed any full size aircraft – Anyway some strange things become apparent with model gliders in strong winds. In 15-20 knots of wind they will go up like kites and you can do all the aerobatics you ever wished for. As well track a slope at right angles to the wind just as you would in a full size glider – At both ends of the slope turn out in to the wind and you can go on and on until you get bored with it – But get too fancy and turn into the slope and things can go wrong pretty quick. Often the second you come off the wind it appears to over power and stall the model and it immediately flicks and spins. This seems to be contrary to the theory advanced by the train carriage analogy. In the same vein I believe I have seen birds similarly overpowered and momentarily losing control before quickly correcting back towards the wind.

    I have always loved anything that flies and have also spent many hours watching birds fly including Albatrosses. Between myself and my son we have taken more than a thousand photo’s of Albatrosses in flight. We have photo’s of Albatrosses in ground effect – so incredibly close to the water they’re only plausible purpose can be to combine the advantage of ground effect with a rising swell. I have also noted that periods spent at low level are often much longer than the time spent pulling up and diving. Hence the reason I have surmised that the majority of the energy employed by Albatrosses and Petrels to maintain continuous gliding flight is derived at low level in ground effect. In which case the pull-up may serve only to look for its prey and/or reposition to a different swell line.

    • Ross whittle says:

      Hi Pete-

      I’ve been flying RC models for almost forty years- I’m testing a new EDF jet tomorrow!!

      The thing about RC is that You are standing on the ground, judging the flight of the model from that frame of reference. As such, many things appear quite different than they would if strapped into the aircraft ( I am also a professional pilot, former competition Aerobatic pilot, experienced skydiver and current glider pilot.)

      Judging the Aircraft’s airspeed by observing it’s GROUNDSPEED is exactley why you will often stall a model when making a turn in a strong wind.

      When you introduce a slope, you completely change the situation as we are no longer dealing with a purerly horizontal flow. What you are describing sounds like surfing in the updraft of the hill, which can extend for a long distance in front of the hill.

      As to a bird deriving energy in ground effect- how? What mechanism is being employed?

  18. Peter Hirst says:

    Half the time this program loses my words!

  19. Peter Hirst says:

    All the time its losing my words now – can’t mess around with it anymore – have to go.

  20. Peter Hirst says:

    Ross I’ll try again to answer your last question regarding ground effect – simply by being physically lifted by rising water and/or slope lift.

  21. Ross Whittle says:

    Hi Pete- That is not what is commonly meant by the term “Ground effect”. Ground effect in it’s usual usage refers to the fact that a wing becomes more efficient when flown at a low altitude above a surface due to it’s wingtip vortices interacting with that surface.

    Any lift derived by from a rising wave or slope lift is simply a case of rising air- yes that would do it, but it is not ground effect.

  22. Peter Hirst says:

    Ross I’m not arguing with you about what Ground Effect is, nor have I been in any of my posts above. For the record I have more than 10,000 hours on 80 different aircraft types and I know exactly what Ground Effect is. I am indicating that I believe Albatrosses spend more time in ground effect than they do pulling up and diving and therefore on balance they would seem to be deriving most energy whilst in Ground Effect with slope lift and/or rising water as the source.

  23. Ross Whittle says:

    That doesn’t necessarily follow- it depends on how much energy they gain in the pull up and how quickly it dissipates during the glide.

    Certainly the most efficient place for the bird to glide is near the surface to take advantage of ground effect, but that doesn’t mean that is where he get’s the energy.

    To give you an example, a good cross-country glider pilot will aim to only spend 20% of his time climbing in thermals, as the goal is to progress on track for as much of the flight as possible.

    He is gaining NO energy for 80% of the time- just seeking to minimise losses-but gains enough in the thermals to “fuel” the glides.

    In a concrete example of Dynamic Soaring, have a look at the modellers breaking speed records with gliders by lea-soaring.

    The only part of their flight that involves gaining energy is the transitions in and out of the lull in the lea of the gill, but it is a sufficiently large “charge” to carry the model trough it’s path, with the craft being efficient enough to still have some of that energy on board the next time it penetrates the shear layer.

  24. Peter Hirst says:

    Ross it’s my strong suspicion that almost all of the energy an Albatross supposedly gains in the pull-up – if any – is actually lost in the plunging glide back to the surface.

    I think the gliders you are talking about have glide angles of 1 in 40 and more. Its unlikely the Albatross has anything like that.

    Regarding you’re last paragraph I am aware that is a commonly referred to explanation, but personally I don’t believe it is anywhere near a complete explanation of that mechanism.

  25. Ross Whittle says:

    Well, the physics is well established and completely satisfies the observed performance, so I’m afraid your personal belief is not backed up by empirical evidence.

  26. Peter Hirst says:

    Ross I also have a strong suspicion that you’ve never seen an Albatross – and that your entire argument above is founded on other peoples work and ideas.

  27. Martyn Johnston says:

    I’ve read about Albatrosses using Dynamic Soaring effects while flying along waves, but I’ve not read of any other birds known to use Dynamic Soaring.

    I write because I’ve just been watching a swallow circling above and below the ridge, on the ‘back side’ of the slope in a manner EXACTLY like model gliders fly when Dynamic Soaring.

    It was blowing at least 40-50mph at the top of Rushup Edge in Derbyshire. The swallow was flying a big big circle, and as he came up above the ridge level he hit the faster air and turned back downwind and went round again. The speed was astonishing. I watched him do at least half a dozen circle before he disappeared.
    He was staying close to the ground all the way round the circle, and his circle was lower, ie more below the ridge than above.

    Is this a known thing that birds do? Has this ever been noticed before ?

    • BOU says:

      Many species frequently make use of different forms of wind and air currents, just as your Swallow was. It sounds like your bird was feeding on the leeward side of the ridge, then at the top used the prevailing wind to take it back to its start point, reducing its energy use in the ‘return’ flight before making another feeding flight.

Leave a Reply

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


£0.000 items