Monday, August 07, 2006

Luneau in 3D

Much argument about the Luneau bird has centered on whether or not any dorsal surfaces of the bird's body or wings are ever visible, and no consensus has been reached. It seems to me a key factor in this is to actually try to determine what the bird's trajectory is through space -- where it is gaining or losing altitude, and how much. Understanding that a bird is neither a missile nor an airplane, its body is not necessarily aligned perfectly with its direction of motion, its wings are not planes, it is still an important clue to know what the bird's motion is relative to the earth's surface and relative to the camera.

Many things make this difficult: the video is shot from a tilted camera in a rocking, moving boat, there is no fixed reference point at great distance, the lack of focus and unknown size of the bird make precise determination of its distance from the camera impossible, etc. But I did the best I could. For a reference I used one of the black blobs in the distant background of the upper left portion of the zoomed in frame. For distance determination I used Cornell's estimate of 20 m as the initial point, and watched how the "white spread" (horizontal spread of the white band along the wings in mid-downstroke, whatever part of the bird this white represents) of the image declined over time. I also used Cornell's number of 30 cm as the diameter of the tree trunk for a calibration. To correct for tilt I looked at the water surface and moss bands on the tree bases where they are visible in the bottom of the frame, and aligned my frame of reference accordingly.

I came up with an estimate of 10-12 m/s as the speed at which the bird is moving away from the camera. This seems like a reasonable speed for a fleeing large woodpecker. I wasn't able to resolve changes in speed over time given the bluriness of the images.

An important point that came from this is that it appears the bird is never really losing altitude. The impression that the bird is dropping initially is an illusion caused by the tilt of the camera and its travel away from the observer. It starts about 1.3 meters above the water, and remains at nearly this same height for the first 0.25 second or so. It then ramps upwards gradually, and in the latter part of the segment before it passes behind the next tree trunk about 1 second later it is climbing upwards at roughly a 10 degree angle. At that point it is about 4 degrees above the invisible horizon, thus its line of travel is tipped upwards towards the camera at a 6 degree angle. At that viewing angle, it seems rather likely to me that you will indeed see some dorsal surfaces on the bird's back and on its secondaries (at least the inner ones) during the downstroke. The bird's angle of ascent significantly exceeds the camera's viewing angle beginning about 0.5 second after it first launches from the tree. While the forward rotation of the distal portions of the wing in flight probably precludes ever seeing the upperside of the outer portions of the wing, the same is not true for the proximal (and trailing) sections of the wing. I think it is unreasonable to assume a priori that the upper side of the bird's inner secondaries could not be visible in these frames. I think it rather likely that they will be visible in some frames. In fact given what appears to be a quite strong forward "twist" to this birds wings in flight, it is quite conceivable that we would simultaneously see dorsal surfaces of the inner secondaries and ventral surfaces of the outer primaries at the same time on the same wing when viewing from slightly above the plane of travel.


Post a Comment

<< Home

Site Meter