Please wait...


Dynamic Soaring

Taylor albatross 4939460473_d1799df1ea

Seabirds such as albatrosses and petrels soar over the Southern and Pacific Oceans using a technique known as dynamic soaring and it is thought that they are somehow using the wind gradient to achieve this. Here is a pilot's perspective on an aspect of bird flight.

Colin Taylor
Retired professional pilot

Avian soaring flight began to be studied in the 19th century and a form of dynamic soaring was first proposed by Lord Rayleigh in an article in Nature in 1883. The debate was initially about observations of pelicans thermal soaring and that kind of soaring has since been explained by other means and practised by glider pilots. The flight of the albatross remained to be explained. It was noted that albatross fly at low level where the wind gradient is found and so the

'Rayleigh cycle' of dynamic soaring appeared to be the answer. Rayleigh asserted that birds cannot soar in a uniform, horizontal wind, and people have been trying to explain dynamic soaring in terms of the wind gradient ever since.

It is assumed that if the bird glides downwind it can gain airspeed from the wind gradient and convert this excess kinetic energy (KE) to height in a zoom-climb. However, if wind-speed is converted to airspeed by transfer of momentum from air to bird most of the energy is going to be lost due to drag which increases with the square of the airspeed. Although the bird can convert excess airspeed to height it cannot, because of drag losses, return to its starting height. Put another way, if the increase in airspeed in the dive is the same as the loss of wind speed in the wind gradient, then KE is the same at the bottom of the dive as at the top. There is no gain of energy, only a more rapid loss of height.

The new Windward Turn theory of dynamic soaring is derived from an analysis of kinetic and potential energies during the kind of manoeuvres that albatross are seen to perform. Here's how it works: In a windward turn at constant airspeed, height energy is gained from wind energy due to a decreasing tailwind/ increasing headwind component. The energy exchange is facilitated by an inertial force, analogous to thrust, derived from the rate of change of momentum due to the reducing ground speed. In a leeward turn, height energy is lost due to a decreasing headwind/ increasing tailwind component. The rate of change of momentum is reversed but losses are minimised by flying the leeward turn in a steeply inclined plane at reduced airspeed. Provided that the bird gains more height energy in the windward turn than it loses in the leeward turn then it can maintain or gain height. See figure 1.
Taylor - ds maneuvre
Airspeed is the same at the beginning and end of the windward turn, but kinetic energy at the beginning of the turn is proportional to airspeed plus tailwind component (squared) and at the end of the turn is proportional to airspeed minus headwind component (squared). The loss of kinetic energy enables a gain of potential energy. As the bird comes around into wind, still at best glide speed (point C), the bird keeps climbing and now reduces airspeed (orange line) and turns the other way. Now at about half best glide speed and a steep angle of bank but at low G and therefore with much reduced drag, it turns across the wind (point D) and dives at some angle to the wind, now with a tailwind component. Having completed the leeward turn at reduced airspeed the bird now trades some height for airspeed (orange line) before commencing the next windward turn.

At the beginning of this leeward turn, kinetic energy is proportional to airspeed minus wind component (squared) and at the end of the turn is proportional to airspeed plus wind component (squared) . At the end of the downwind turn, kinetic energy is greater and therefore height (potential energy) is lost. The mechanism of transfer of momentum is the reverse of that in the windward turn.

The bird can maintain height using dynamic soaring but can only do so at the expense of drifting downwind. This is probably why albatross circumnavigate the Antarctic continent with the prevailing wind. Provided that the energy transfer in the windward turn, is at least equal to the rate of loss of energy due to drag then the bird will maintain airspeed and maintain or gain height but only until the turn is complete.

The Windward Turn theory shows how dynamic soaring can work but contradicts most previous work on the subject. It shows that there is a possible mechanism to exploit a uniform horizontal wind at any altitude and there is no need to invoke the wind gradient or high G manoeuvring or high airspeeds. It also shows that there is not enough kinetic energy in the wind gradient to account for dynamic soaring, that although dynamic soaring happens in the same part of the sky as the wind gradient, the wind gradient actually reduces the efficiency of the manoeuvre and that dynamic soaring probably works better over the oceans than over the land because the wind gradient is smaller and the wind is stronger and steadier. It also helps to explain some aspects of bird physiology such as tube nostrils and high aspect ratio wings and of behaviour such as the direction of large scale flight patterns and the shape of small scale repetitive turns and also why the bird's metabolic rate is low whilst gliding, because the manoeuvre is carried out at relatively low G. There is a more detail in the website.


Related links

Dynamic Soaring for Birds

Image credit: Black-footed Albatross © Bill Boulton

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


Blog with #theBOUblog

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

28 comments on “Dynamic Soaring

  1. Avatar Colin Law says:

    If I understand this blog correctly it suggests that a gliding bird could retain height without any effort, in steady wind, at the cost of drifting downwind. This would appear to contravene the conservation of energy laws. Also presumably a full size glider could use the same technique to stay aloft indefinitely.

    • Avatar BOU says:

      Colin Taylor replies –

      You are right that the theory says a glider can maintain height in a uniform horizontal wind, provided the wind is strong enough. That is what the albatross do. However, far from contravening any physical laws, the Windward Turn theory is derived directly from the principle of conservation of energy and Newton’s laws of motion and momentum. It shows that the kinetic energy of the wind is converted to potential energy in the height of the bird and in overcoming drag. The energy comes from the wind and not the wind gradient.

  2. Avatar Colin Law says:

    I think that the suggestion that a glider could maintain height indefinitely when there is no thermal activity, even in a strong wind, with no shear, will come as a surprise to glider pilots. We need some input from someone with some qualifications to make a definitive statement on the fact that the whole premise of this blog is incorrect. To have something like this on the BOU website, which would appear to give it some authority, is not appropriate.

    • Avatar BOU says:

      The intro to this blog post clearly states this is the perspective, and therefore views, of the author. One of the roles of BOU is to facilitate discussion and debate, so in this respect we consider it appropriate to host items such as this so that those with an interest and/or expertise in the subject can exchange views..

  3. Avatar Colin Law says:

    The author states that a glider, or bird gliding without flapping, could maintain height in a steady strong wind using the Windward Turn theory that he describes.

    The first point to note is that once a glider (or any flying object for that matter) is away from the ground the only forces acting on it are gravity and the effects of the air flowing over it. The way a plane flies is not affected in any way by the ground speed, only by the wind speed over and around it.

    The implication of this is that the forces on a plane flying at, say, 50mph through the air are exactly the same whether there is no wind, a tail wind of 50mph or a head wind of 50mph. This will affect the speed over the ground obviously, but not the way the plane flies through the air. In fact as any pilot will tell you, if the ground is not visible it is impossible to tell which direction, or at what speed, the plane is moving over the ground.

    This implies that if the authors suggestion that, using the Windward Turn theory, a glider could maintain height in a steady strong wind were true then using the same flight pattern (through the air that is) would allow a glider to maintain height in calm conditions, which is obviously not correct.

    Finally for anyone who does not believe that Dynamic Soaring works have a look here. This shows a radio controlled model glider using the technique. In that clip the wind is blowing from left to right over a ridge and the glider is using the wind shear between the strong wind blowing left to right above the ridge and the dead air below the ridge on the right hand side to stay aloft. If one imagines the ridge to be a wave, which would be moving from left to right, and the model glider to be an albatross then I think one gets the idea. A large hill is not necessary, I have seen this done on the back of a sand dune, and if a model glider can do it on a sand dune then an albatross with sensors at the base of every feather telling it what the wind is doing, can certainly do it on any decent sized wave.

  4. Avatar Adrian Chapmanlaw says:

    Having been an RC glider pilot for many years mainly slope soaring, this “theory” confuses me. I have flown in wind conditions from 3mph to 70mph and the thought of zero gradient height retention is just impossible on a slope as soon as the ground levels out the lift disappears. hovered a model into my hands in a wind on many many occasions.

    I have also done a lot of dynamic soaring using a spine back ridge to create a heavy wind gradient. I have flown speeds well in excesses of what the model is capable of in a terminal velocity dive, never mind a circle pulling inexcess of 30g. I have personally flown at over 150mph and in the USA where high altitude mountains are used they are seeing speeds over close to 500mph. (youtube go look at youtube

    It is simply not possible for DS to not be a result of wind gradient. The sound of a model entering the wind shadow of a hill, the audible increase in airspeed that comes with it and the sound the model makes which at times you can feel in your chest as it crosses the boundary layer back out into the airflow.

    physics blows this theory out of the water… however the physical evidence from the rc gliding community throws a nuke into the mix.

  5. Avatar BOU says:

    Colin Taylor replies –

    The points raised by these corespondents are all covered in the website.

    The result of this theoretical research is surprising but the mathematical analysis speaks for itself. It is equally surprising that the wind gradient theory has persisted so long when it is so full of holes. In 2006 Nasa tried to test the wind gradient theory with a Blanik glider and got nowhere with it. Ironically their results tend to favour the windward turn theory. It should not be surprising that, occasionally, old (and untested) theories are found to be wrong and new theories take their place.

    It is ludicrous to suggest that flying the dynamic soaring manoeuvre in still air will have the same effect as flying in a wind and that is not what the article says. It is a fact that windspeed affects ground speed and therefore affects momentum and produces inertial forces. That is my personal experience after many years of flying.

    The model glider flying is obviously not the same dynamic soaring as practised by the albatross. It is, in my opinion, more like a kind of autorotation as explained in the website.

  6. Avatar Colin Law says:

    I did not say that windspeed does not affect ground speed. What I said was “The way a plane flies is not affected in any way by the ground speed, only by the wind speed over and around it.” Can we agree on that?

    • Avatar BOU says:

      Colin Taylor replies –

      No we cannot agree on that. An aircraft is affected by all of the forces on it including inertial forces which are affected by momentum, by groundspeed and by the wind. Thats why birds land facing the wind – to reduce their momentum and kinetic energy before they hit the ground.

      • Avatar eyytee says:

        “Inertial forces” exisit only in non-inertial frames. In the rest frame of the Earth there are no relevant inertial forces actin on a bird. Groundspeed is relevant during landing, but not during flight.

      • Avatar windgrins says:

        The dynamic soaring of an albatross is wonderful but it doesn’t occur without changes in the velocity of the wind in which the albatross is operating.

        Unfortunately the author of this article (in spite of being a retired professional pilot) is confused about kinetic energy and reference frames. There is no way, without adding energy to the system, for a bird or airplane to increase its altitude in steady wind once stabilized, without a loss of airspeed. In fact, over time, in a truly steady wind or calm air, without input energy (engine or flapping), the plane or bird will lose altitude due to drag losses. A steady wind cannot supply this energy.

        The comments about kinetic energy confuse the inertial reference frame of a steady wind with that of the ground. Gradients of some sort must be present for dynamic soaring to be feasible. Since gradients produce a wind velocity at different altitudes or locations, the bird’s inertia can be utilized to convert the energy in these differing winds to altitude which makes dynamic soaring possible. However this has nothing at all to do with ground speed. The author is correct that airplanes and birds land into the wind to reduce kinetic energy over the ground as the two media (ground and air) are moving relative to one another. Once flying at a specific altitude and airspeed, neither an airplane nor a bird can exploit a truly steady wind to increase its altitude without adding energy. Once the bird is in smooth air with no gradients, it’s speed over ground has no effect on the birds flight as was mentioned by earlier posters contrary to the authors assertions here. The author’s calculations of kinetic energy, confuse the inertial reference frame of the ground and the steady wind and produce an incorrect answer and flawed explanations.

        Sorry, but the basis of this article is simply incorrect as it is a violation of conservation of energy and his perceptions and physics are flawed.

  7. Avatar eyytee says:

    Colin Taylor: “It shows that there is a possible mechanism to exploit a uniform horizontal wind at any altitude”

    This would violate basic physics, like Galilean invariance. The rest frame of an airmass moving uniformly relative to the ground is equivalent to the rest frame of an airmass at rest relative to the ground.

  8. Avatar Ross whittle says:

    The mistake in this thinking is illustrated in Mr Taylor’s last response.

    When in flight, an aircraft or bird is in no way effected by it’s ground speed. Mr Taylor makes the mistake of using the Earth as an absolute frame of reference, thinking that any gain in ground speed is gain in energy.

    Kinetic Energy is a RELATIVE value- it varies depending on frame fbreference. An aircraft circling in a constant wind may gain and lose energy with reference to the ground, but that has no bearing on it in flight.

    From a frame of reference at the same speed as the wind ( which is what the aircraft or bird is interacting with) there is no variation in energy, and thus none to be exploited.

  9. Avatar BOU says:

    Colin Taylor replies –

    These responses are very disappointing. All of their points are covered in detail in the dynamic-soaring-for-birds website. Anybody reading it will find a reasoned argument supported by mathematical analysis which is derived directly from the principles of conservation of energy and NewtonÔÇÖs Laws of motion. Simply stating the contrary does not constitute a valid counter-argument. The Windward-turn mathematical model of dynamic soaring accurately predicts the wind strength needed for albatross to fly. Simply stating that it gives the wrong answer is again not a valid counter-argument.

    The website explains, amongst other things, that the bird exists in two inertial frames of reference: one attached to the moving airmass against which airspeed and drag are measured and one attached to the surface against which the wind and the birds groundspeed and the birds energy and momentum are measured. This is fundamental to the navigational triangle of velocities.

    When I began studying ornithology I was quite prepared to accept the wind-gradient theory of dynamic soaring but I wanted to understand how it worked, so I began to develop a simple mathematical model to do that and account for drag losses. I soon discovered that the wind-gradient theory simply does not work in any practical way. The windward-turn theory came out of that analysis.

    I stand by everything in the blog and the website. I am not making anything up. The mathematics speaks for itself.

    The only thing I am confused about is why people are so keen to cling to the wind-gradient theory which was invented in 1883, before people had any experience of gliding flight.

    Colin Taylor

    • Avatar Ross Whittle says:

      But Colin,

      Birds do no “Exist in two frames of reference”. ALL things exist in ALL frames of reference.

      What you can’t do is work out energy in in F of R, then try an liberate it in another. THAT vilotes C of E.

      What velocity anything has WRT the ground has NO effect on it’s relationship with the Air. Yes it effect navigation, and yes we takeoff and land into the wind, because in thos cases, we WANT to interact with the ground.

      But that makes no difference to the aerodynamics of an object WRT the air.

      For instance, I’m sure you agree that the IAS for Vs does not vary with a head or tail wind?

      You must make all calculations with reference to one F of R, or you ARE breaking basic Newtonian physics.

      Colin, I HAVE extensive experience of Glider flying, Parachuting and Aerobatic Flying, as well as being a Jet Captain/TRE. You CAN’T gain energy by utilising a constant horizontal wind.

      I’m sorry, but you are simply wrong here.

  10. Avatar Ross Whittle says:

    To expand a little further- Momentum is NOT measured wrt the Earth. It is a relative value measured against whatever frame of reference you are using.

    To say a birds momentum changes relative to the ground because of wind, then suggest this can be used to gain height by using it to increase lift relative to the air, is to switch Frames of reference mid calculation, which is not a valid calculation in Physics.

  11. Avatar windgrins says:


    I read many of the items on your website and noticed there were factual errors as the posters above describe. For example, you discussed the idea of an aircraft in a constant power, constant speed 360 degree turn while maneuvering in a constant wind. You state “So, when the aircraft turns from tailwind to headwind, ground speed reduces, momentum reduces, there is a negative acceleration and hence a positive or forward inertial force. That inertial force is analogous to thrust and tends to make the aircraft climb. When the aircraft turns from headwind to tailwind, ground speed increases, momentum increases, there is a positive acceleration and a negative inertial force. This inertial force is analogous to airbrakes and tends to make the aircraft descend. All at constant airspeed.”

    This is simply wrong. There is no net force change along the longitudinal axis as the aircraft executes a simple circle in the reference frame of the wind but a different pattern over the ground. As the poster above states, Kinetic Energy and momentum are relative values to the frame of reference and not only to the earth and in fact simultaneously have different values when computed in the two reference frames. The earth/ground does not constitute a “preferred frame” of reference, just a convenient one for many situations. Even though, the values for KE and momentum take on different values at a point in time in the two frames, All conservation laws are preserved in any inertial frame at all times.

    The reason the posters hold these principles dear is not to preserve the concept of dynamic soaring in a gradient situation but to maintain the long established concept of Conservation of Energy which your assertion violates. The math is not what is at issue, your physics is. Birds nor airplanes can neither increase altitude with no change in airspeed in either calm conditions or truly steady winds without a source of added energy (motor, flapping, etc.). From a physics standpoint, both of these situations are the same and the two situations constitute a (near) inertial reference frame and all physics proven in one holds in the other. Unless there is interaction by the bird/vehicle with both media (air/earth), there is not a mechanism to change its energy and the altitude will decay due to drag over time.

    You are welcome to discuss these ideas with a group of individuals who have a lively discussion about these exact issues. There is a thread about the subject of “Directly Down Wind Faster Than The Wind-DDWFTTW” that is currently in progress at TalkRational.org. Your article was brought up on the web site for discussion there. Many at that site will be happy to explain the fallacy in your assertion. Please join us there.


  12. Avatar BOU says:

    Colin Taylor replies –

    Ross Whittle

    The point I am making is that airspeed is measured relative to the air and groundspeed is measured relative to the ground.

    The energy I am concerned with is the KE and PE of the bird relative to the ground. I am taking the birds inertial speed (total speed or ground speed if you like) and measuring the difference between KE with a tailwind and KE with a headwind. The difference in KE is available to convert to PE, which is then available to gain height or overcome drag.

    I am not saying that groundspeed affects aerodynamic effects. I am saying that changes to ground speed during turns cause inertial forces which act on the mass of the bird. That after all, is how aircraft inertial navigation systems work.

    Effects can transfer from one frame of reference to another. For example when an aircraft turns due to aerodynamic forces relative to the air the pilot experiences inertial forces (G loading) relative to the aircraft.

    I have described theoretically that you can gain height in a uniform horizontal wind. Simply denying it, however politely, does not constitute a counter-argument.

    If momentum is a relative value then I choose to measure momentum relative to the surface.

    I am not saying that changing momentum increases lift. I am saying that changing (reducing) momentum produces an inertial force in the direction of flight. That force is analogous to thrust and tends to increase airspeed or height.


    Again denial does not constitute a counter argument. The paragraph you quote is a valid analysis of the situation described and reflects my personal experience of circling in light aircraft. The effect is small but the wind does have an effect and generally makes a circling aircraft lose height.

    Force causes acceleration and acceleration causes inertial force. Acceleration of groundspeed due to changing wind-components when turning an aircraft causes inertial forces to act on the mass of the aircraft.

    The surface is my preferred frame of reference because the critical value is the height of the glider above the surface, therefore I am comparing KE and PE relative to the surface. Momentum is also measured relative to the surface but when momentum changes, the inertial force acts on the mass of the bird. That force then also exists in the air-borne frame of reference and acts like thrust or airbrakes.

    My assertions do not violate the conservation of energy, they are based upon the application of that principle.

    The effect of inertia makes turning an aircraft in a wind different to turning in still-air. Considering this, the windward-turn theory is the mechanism that best describes dynamic soaring. It explains what the albatross does and how it does it. It explains aspects of albatross physiology and behaviour. The wind gradient theory does none of these things. Is there anyone out there interested in learning something new?

  13. Avatar Ross whittle says:


    We have not simply asserted these things, we have explained why your ideas are wrong.

    HOW is the change in KE relative to the ground able to be used to increase the birds altitude UNLESS it is able to use it to increase lift?

    Are YOU interested in considering you ideas might be wrong, as they clearly violate Newtonian Physics?

  14. Avatar Ross whittle says:

    As to the surface being you preferred frame of reference, energy calculations must work in ALL frames of reference.

    If you see that you cannot calculate a KE change the leads to the ability to climb if you reference a frame at the speed of the wind, you have PROVED it cannot work, as it MUSY work in all frames of reference.

    Your idea that the only frame of reference that counts is the surface because that is what height is measured against is wrong- you are ALSO climbing relative to the frame at wind speed.

    Try it. Do your calculations ONLY referencing the birds airspeed. If your calculations don’t work, you’ve shown your ideas are erroneous.

    Please go to the Physics Forums, register and put forward your ideas. The site is inhabited by PHDs and practicing physics professors. You talk about learning something new- are YOU willing to do so?

  15. Avatar Ross whittle says:

    And I have turned many, many aircraft in VERY strong winds- the aircraft has no idea what it is doing reference the ground, and turns EXACTLEY as it would in nil wind.

  16. Avatar BOU says:

    Colin Taylor replies:

    Ross Whittle

    Please report to your nearest flying school. You do not gain altitude by increasing lift. Lift is always approximately equal to weight. You gain altitude by increasing thrust. In the case of dynamic soaring there is an inertial force analogous to thrust.

    I have analysed dynamic soaring according to the wind gradient theory and found it to be wrong. Out of that analysis came the windward turn theory which complies with the principle of conservation of energy and Newtons laws of motion.

    You only have to consider relevant frames of reference. If airspeed is constant then there is no change of kinetic energy in the air-frame of reference but if ground speed is changing then there is a change of kinetic energy in the ground-based frame of reference.

    In a horizontal wind a gain of height above the ground is obviously also a gain of height in the air. The gain of potential energy is due to reducing kinetic energy due to changing ground speed.

    I have done the calculations referencing airspeed, they are on the website. They prove that the wind-gradient theory is wrong. The windward turn theory works because airspeed is constant during the windward turn.

    An aircraft turning in a wind does not turn exactly as it does in nil wind. In still air the ground speed is constant. In a wind, groundspeed changes, which leads to changes in kinetic energy, momentum and inertial forces.

    May I point out that I am the one with the new idea.

  17. Avatar BOU says:

    This correspondence is now closed.

  18. Avatar Aravind says:

    It should not be difficult to simulate the strong and steady wind that’d be necessary to maintain height in a flight simulation software.Am I wrong?

  19. Avatar Rick Cavallaro says:

    Unfortunately Mr. Taylor’s analysis is simply hopelessly wrong. Dynamic soaring is very well understood and very well described by many people with expertise in both real world flying and theory. It’s well known that an object operating in an inertial frame will behave exactly as it would in any other inertial frame. In other words, an unpowered aircraft could no more maintain altitude in a steady with without gradient than it could on a day with no wind.

    Just as a sailboat harnesses the energy of the wind relative to the water, a glider harnesses the energy of the faster wind relative to the slower wind in the gradient while dynamically soaring.

    I strongly encourage anyone with questions about this to refer to most any other article on the topic.

  20. Avatar Llyricist says:

    There are no fictional forces in an inertial frame of reference, much less one that is analogous to thrust.

    The fictional forces that do exist in a frame of reference fixed to the earth’s surface are too negligible to make any significant difference at the speeds, distances and times under discussion.

    The centrifugal and centripetal forces felt by a person inside a turning aircraft are not felt by the aircraft in a frame of reference fixed to the earth’s surface, or even one fixed to a homogeneous air mass in motion with respect to the surface.

    The significant forces acting on the bird not providing thrust in those frames are all very real, and limited to lift, drag and weight.

    The only propulsive force is the longitudinal component of weight. Which only exists if the bird is descending. In order to initiate a climb, the bird has to increase lift, which increases drag, which reduces airspeed, which reduces lift. To turn, the bird must increase lift, which increases drag, which reduces airspeed unless the descent angle is increased to compensate.

    To both climb and turn, the bird needs to provide thrust.

    This of course, only applies to a magically homogeneous air mass close to the surface, in the real world, the air mass will have a wind gradient, eddies and vertical components of velocity which have very real significant effects.

  21. Avatar Steve Seibel says:

    See a related discussion of the ideas on Mr. Taylor’s website here —


    Sincerely, Steve Seibel

  22. Avatar Steve Seibel says:

    Discussions of Colin Taylor’s ideas have been ongoing on the web forum https://www.rcgroups.com/forums/showthread.php?2910873-Alternative-facts-oddball-theory-of-dynamic-soaring/page15#post39573248. See posts #215 and #216 for a comprehensive analysis of the problems with Mr. Taylor’s theories.

    Sincerely, Steve Seibel

Leave a Reply

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


£0.000 items