As treatment processes are kinetic-dependent, a consistent description of water residence times is essential to the prediction of waste stabilization ponds performance. A physically-based 3D transient CFD model simulating the water velocity, temperature and concentration fields as a function of all influent meteorological factors – wind speed and direction, solar radiation, air temperature and relative humidity – was used to identify the relationships between the meteorological conditions and the hydrodynamic patterns and water residence times distributions in a polishing pond. The required meteorological data were recorded on site and water temperatures recorded at 10 sampling sites for 141 days. Stratification events appear on very calm days for wind speeds lower than 3 m s−1 and on sunny days for wind speeds lower than 5 m s−1. De-stratification is related to two mixing processes: nightly convection cells and global mixing patterns. Numerical tracer experiments show that the results of the flow patterns can be evaluated using the dispersed flow regime approximation and, for wind speeds exceeding 6 m s−1, the completely stirred tank reactor assumption.