Major Tom:
Just to clarify: the acceleration range I quoted is for ALL the data I have looked at and involves at least TWO different videos. The precision on any ONE given plot is, of course, much better.
The main problems in getting precise numbers are:
1. Defining t(zero), the moment the roofline starts to move downward.
2. Measuring each drop distance. This is most difficult during the first second of motion. Here an uncertainty of 1 meter is as large as 30 % of the drop. After that the drop error gets progressively smaller. Thus at about 4.5 seconds the building has dropped about 86 meters. The uncertainty in the drop distance measurement might now be 2 meters, but this is only about 2 % of the drop.
