Some calculations fr Smolensk forum.
Standard Russian glissade - 2.66 (2 degrees 40 minutes). And theoretically, the TU in Nortern should have followed this route:
Far Beacon distance 6,100m, height 300m.
Near Beacon 1,100m, height 70m.
Let's count:
tan(2.66/180*pi)*6100 = 283.4008 (edge of the runway, height 17m)
tan(2.66/180*pi)*(6100-1100) = 232.2957 (when above Near Beacon, height 68-70m.)
(Checked by Windowds in-built computation ? device
2.66° tg * 6100= 283,40076306540161906494420247723 m)
Now, an Aml idea. "What if the navigator, instead of Russian 2.66 took IKAO recommended 3 degree glissade? Then, to the edge of the glissade the plane arrive at -20m; nearly 40 m difference to what it should have been. While they were going over the ravine all was compensated more or less. But then there started the climb up of the terrain and I think at approx. 1,200m from the edge of the runway the systems warned them of the dangeous nearing the surface. At 1,200m away from the runway, lowering at 3 degrees, they'd have height 46m from runway level - at the planned height of 72m.
On finding this, the crew changed regimes for take-off at which point the aircraft sank even lower. An idea to think over."