Against the backdrop of a lack of fidelity of conventional deterministic parameterization, more physically based stochastic parameterization schemes have been implemented in weather, seasonal and climate models. Stochastic approaches for representing multi-scale cloud and convection parameterization, have shown particular promise. Although stochastic approaches are advantageous in breaking the quasi-equilibrium deadlock, the time has come to take a holistic look at the ability of stochastic parameterization to capture the spatio-temporal variability of processes such as clouds and precipitation and their linkage to climate, Sub-seasonal to Seasonal (S2S) and weather variability. One aspect of a stochastic parameterization worth exploring could be its fidelity in capturing weather and climate extremes. This session is planned for a detailed discussion on stochastic parameterization of physical processes and also would provide future pathways for improving the sub-grid scale variability of processes such as clouds and precipitation in global climate models across the scales.