Estimating Rollout Speed from Videos
Having 2 video views of the aircraft rolling out allows estimating the speed of the aircraft based on a number of measurements (assumptions) - first order estimates at the very best. Of course the final accurate answer will hopefully come from the recovered FDR. I offer this as just an exercise and NOT in any way a definitive estimate of speed of the accident aircraft.
From the end view video, it appears the aircraft lights dip noticeably at about 22 seconds (post # 51 video) and collides with the camera assumed to be mounted on the ILS localizer tower at 27 seconds. The map view of the runway shows a culvert just past the runway, likely draining into the river. Assuming the apparent dip of the aircraft (main gear still on the ground) was caused by rolling over this depression, and knowing the time interval AND distance between the culvert and ILS tower would give the speed of the aircraft. From the scale on the map I measure 413 meters (+/- ???). Traveling 413 meters in 5 seconds would give an 'average interval' speed of 161 knots. NB The time period (5 seconds) is the weakest assumption and likely largest source of error here.
From the side view video what's needed is the frame rate of the camera (frames per second). The scale could be determined by using the published length of the aircraft (YAK-42D ??? 36.4 meters) and measure the displacement frame to frame of the aircraft measured in length units of the plane on the screen. For example if the frame rate is 4 per second (0.25 second) a 36.4 meter long aircraft traveling at twice its length per second (72.8m/sec = 142 knots) should be displaced by half its length per frame.
What's needed is better video processing software with better time resolution on both videos - I don't own any, so I make no estimate here. Someone who does with some time to kill, might want to carry this exercise further with possibly a bit more accuracy. Use the last 2 or 3 frames as the aircraft goes off screen left because the camera view there is more perpendicular to the runway - less perspective error.
Again I offer this as only an excersie or a template for analysis and make no claim of accuracy for the limitations I stated. We of course await the actual numbers from MAK and the FDR.
Thanks.