Originally Posted by
Intruder
For many things, they can only certify what they test to.
Not quite - you're allowed to extrapolate to some extent, but generally allowed when there is some previous data that validates the extrapolation methodology.
Another issue is many of the data tables only extend over a specific temperature and altitude range - e.g. engine ratings. I recall an issue perhaps six or seven winters ago when a big cold front hit part of Europe - it was bitter cold, but more importantly the pressure altitude dropped well below -1000 ft. The takeoff performance data for some types didn't go below -1,000 ft., resulting in several aircraft being grounded for several days.