ALT Example 1: A) 50%:50% liquid-water:ice in dynamical coexistence in correspondece of the melting T. B) LENS (similar to, e.g., local diffusivity) time-series data for all molecules in the system. C-F) Iterative detection of evident (most populated) dynamical environments/domains. At the time-resolution of, e.g., dt=0.4 ns, Onion Clustering detects 4 statistically-relevant domains & determines their thresholds (D,F) - black: bulk ice, red: ice-water interface, cyan: liquid water, blue: local freezing/melting ice-like domains in the liquid phase. E) In yellow the ENV0 cluster: the amount of data that cannot be classified as statistically relevant (events that occur faster than dt=0.4 ns). G) Number of statistically-relevant clusters as a function of the time-resolution dt (in blue), and fraction of "undetermined information" (in orange): above dt=10 ns, the resolution is too low and fraction of ENV0 data increases. H) reconstruction of the internal microscopic dynamics of the system.