OK, most of the explanations posted are interesting, but totally wrong (and many not even possible). Here’s the final conclusion:

Like Gergely Vass said, it was done with something like a pushbroom scanner type of technology. My very first video digitizer from way back in ‘84 or so did this type of distortion on moving objects. Although, mine scanned from left to right from a video image coming in, taking 6 seconds for a single image to come through and if the target object was moving during the digitization, it would be warped.

Here’s the difinitive answer to the warped plane:

What it is NOT:
1. The plane is NOT banking.
2. The distortion has nothing to do with it being near the edge of the photo.
3. The distortion has nothing to do with it being higher than the ground beneath it.
4. The distortion has nothing to do with the angle of the satelite to the ground.
5. The doppler effect most definitely has nothing to do with it… that can only effect the wavelength (for light, that’d be color)… and for any moving object to have a noticable doppler effect with visible light, it’d have to be traveling at relativistic speeds (a large portion of the speed of light), which a plane moving slower than the speed of sound will most certainly NOT do. No object the size of a plane can be pushed to such speeds by current technology, and certainly not in our atmosphere or it would burn up and break apart and would be moving too fast to be captured in a photograph.

What it IS:
1. It is a side effect of:
a. The motion and speed of the plane relative to the relatively fixed aiming of the camera to the ground.
b. The slowness of the scanning process of the device capturing the image.
c. The direction of the scan (top to bottom in this case).
2. The device that captured the image has some kind of a video camera.
3. The video image is captured one horizontal line at a time.
4. Each line that’s lower in the image is from a later point in time that the line above it.
5. While the image capture device was capturing horizontal lines for this image, the plane was moving down and to the right, resulting in each horizontal line showing the plane at a slightly different position further down and further to the right.
6. This resulted in an image with the image of the plane skewed in the horizontal direction it’s traveling (to the right from top to bottom of the image) and vertically (causing the plane to appear taller).

If the plane had been moving upward, it would have been squashed vertically, rather than stretched vertically. If the plane had been moving to the left, the horizontal skew would have been towards the left (from top to bottom) instead of to the right.

To prove this, I’ve done an incredibly simple image manipulation of loading the image into a simple page layout program (InkScape (free software from http://www.inkscape.org/) but just about any image manipulation software can do this). I then distorted the image by skewing the bottom towards the left. Then, I shortened the image vertically.

I’ve posted the final image here:

http://www.easiesttoremember.com/warped%20plane%20resolved.jpg

As you can clearly see, the plane is normal, but the distortion has been transfered to the background.

My skewing to the left compensated for the right to left movement of the plane in the timespan of the image capture. My vertical squishing compensated for the downward movement of the plane during the timespan of the image capture.

I hope this puts all the questions to rest.

The halo effect is common for objects on the edge of a lens - chromatic abberation.

I recognized the location immediately because I often fly this rout.
It has nothing to do with SFO.

The planes landing at San Jose Airport (SJC) makes their turns right here.
If you are taking pictures from land you will see the airplanes tipped in a hard angle, and the photos will look like as if you took a picture flying right beside it.
As to the blur, it could be a reflection of light and of course, movement.