Supercooled water is a metastable thermodynamic state and, therefore, the associated phase transition (to ice) must be irreversible. Has this irreversibility been properly addressed? Does it really matter to an atmospheric scientist? I'll argue No and Yes, respectively.

We use measured temperature-dependent heat capacities of supercooled water and ice to calculate the ice-(metastable) water entropy difference and to estimate a lower bound on the amount of latent heat, liberated by the freezing droplets. The calculation is compared with tabulated values of the latent heat of fusion with surprising results. Based on a novel physical
picture of the freezing process, we suggest a simple estimate for the effective latent heat, suitable for heat budget calculations of glaciating clouds. In addition, we arrive at a quadratic dependence on supercooling for the irreversible contribution to heat exchange during the freezing process. Implications for optical properties of the "hurriedly made ice" will also be discussed.