Girma Orssengo rightly demonstrates that one can not determine climate sensitivity empirically from observed changes in CO2 concentration and in global
mean surface temperature unless one either studies periods that are multiples of ~ 60 years to cancel the transient effects of the warming and cooling phases of the Pacific and related ocean oscillations or studies periods centered on a phase - transition in the ocean oscillations.
The air
temperature in the troposphere is governed by the
surface, so it isn't going to warm first anyway
unless you do things like insulate it from the ground which
means it is not physically the troposphere anymore.