It is fairly normal: the efficiency of the whole process (pump coupling, solid state laser and doubling) all depend on temperature.
One thing often overlooked is the first part: The 808 nm pump light that powers the solid state laser has to be within a rather narrow range of wavelength. The actual wavelength produced does however depend on the pump diodes temperature. So does the range that is acceptable as pump light for the solid state laser, albeit to a lesser degree.
If this is off for any reason, the pump energy will not fully be absorbed by the solid state laser and output power will drop. It may start working better again once temperature has evened out across the whole laser.
The complexity of all these steps and the narrow tolerances can sometimes lead to strange situations. Often 532 lasers as a complete unit have some optimal working temperature, but in some cases there might be more than one. You could, for example, have one that works well at 25 degrees and also at 40 degrees, but not that well at 35 degrees (centigrade). If this is bothersome in your application you could consider getting a replacement unit, but all of them have a temperature optimum unless the unit is thermally stabilized.
