Several reasons, the most important is that there are in fact very few reliable standards for temperature calibration performed at cooling. Secondly, certain processes exhibit significantly less pronounced effects (wide and shallow peaks in case of reversed themally activated processes; simple step change at Tg) during cooling. Thirdly, most of the methods of kinetic analysis used to evaluate the kinetic parameters of the studied processes are derived only for the heating or isothermal conditions.

Thank you Roman Svoboda