"Estimated" component data which KCAS values are based on, starting at 9:03 (represented by the light green line of the TSB flight data chart above dark-green/groundspeed) is pretty close. Maybe is a bit faster AS at a couple of spots ...Sidebar wrote:I think the TSB probably used "estimated" because they were using met data from an observation drone about two hours before the accident.
The "radiosonde" component-info (chevrons between 2500' and somewhere just below 5000') are given North at 20-25kts (looks like) for/after the early altitude ("5kts-NNE"/lowest chevron) where the northerly course started around 2500ft around the left turnout. IMO there must be at least the 20kt spread between these two speedlines for the middle of this five minute northerly climb .. in-between 9:03:10 and 9:08:30.
The first track data "9:03 to 9:03:34" includes "800fpm" for 10-15sec and the "100ft" lost right-after, which is signature of some brief performance increase, yet still ends up only 240ft of net altitude gained in that 34sec segment.
Following the +/- fluctuation and after another five minutes of climb to 9:08:41 there is obviously performance decrease (shearlike) where the Cargo Caravan transitioned into lighter/warmer West component moving at/above 5000' (indicated by radiosonde-infochart/wind-chevrons) near where the climb-rate ceased. Such relativewind reduction going-on ahead-of "event 1" is a secondary climb-performance loss coming alongside the primary one (the lift-degrading icing buildup/encounter described in the report).