I am having trouble figuring out the logic behind the way that obstacle clearance areas are constructed. Take, for example, a departure procedure where Initial Climb Area is defined in TERPS (section 14-1-6) as a trapezoid:
Even as far as subsequent segments are concerned, their respective obstacle clearance areas seem to follow the same pattern, like this example from TERPS:
Why is this happening? Is there a reason that at the beginning of each segment the obstacle clearance area "shrinks" and then splays outward? Especially when one considers SIDs with really lengthy segments, where near their end the aircraft's perpendicular distance from OCA's borders is huge. It would seem more reasonable to me if the obstacle clearance area was a rectangular region, as it is the case with STARs:
Could someone point me in the right direction?