dT=dt/(M*c)*(Pcav*a+Pcoils)+dTamb
where dt is the time interval, M is the mirror mass, c is the specific heat, Pcav is the light power in the Fabry-Perot cavity, a is the mean coating absorption. In the formula is shown also the warming up due to the RM coils and the ambient temperature fluctuation. In the following evaluation the Pcoil is neglected, since no apparent difference in the coil signal is visible for the WE coils with respect to the NE (see third figure).
In the fourth (WE) and fifth (NE) figures a zoom of a lock-delock sequence is shown and the larger temperature increment of WE is evident. This seems not justified by a change of the ambient temperature, as shown in the sixth figure, where the temperatures measured on F7 are shown(and the fluctuations are less than 1/20 of the fluctuations measured on the mirrors).
Hence, if you believe to all that and if we consider the same power (about 13kW) in the North and West cavities it is possible to state that the difference in warming up of the mirrors is driven by a difference in absorption in the coating, with a larger absorption in the WE. An attempt to measure the WE absorption is shown in the last figure, for a series of lock-delock. In that figure also a partial and incomplete evaluation of the measurement error is shown. The measurement errors are still too large to make any clear statement, but attention in the future should be given to the WE behavior.
