Thanks to sheer luck, two people in the Denver area captured a photo of the same lightning bolt from very different locations. Richard Wheeler to decided to try and use those two images to reconstruct the lightning bolt in Google Earth and seems to have done a pretty good job!
Richard went through a number of steps to create the model of the bolt:
- Scaled both images to the same size
- Traced both images and matched up the coordinates of each location in the bolt
- Put the resulting data in a table to calculate the difference in x and y position in each image
After he had done that, he had some work to do:
Now we need to do some maths… except I don’t like doing complicated maths and it turns out there is a big simplification you can make! If both pictures are taken from a long way away from the lightning bolt (i.e. the object has quite a small angular size in the image) then the shift in position between the images is proportional to the distance from the camera. Bigger shifts mean that bit of the bolt is closer to the camera. This approximation is pretty accurate for the majority of cameras, so I used it here.
The other problem is the proportionality factor. If one part of the lightning bolt shifts twice as much between the two images as another part that means it is twice as close. But twice as close as what? Without knowing exactly where the cameras were positioned that means only the relative distance, not absolute distance, can be calculated. Oh well, close enough!
You can view the resulting image in Google Earth by using this KMZ file, or read more about the process of creating the file on his blog.
Great work Richard!