The improvement of numerical models for predicting weather and climate at variety of space and time scales have taken place globally. However, there are still area of challenges which need to be addressed with better understanding of physical processes based on observations and subsequently representation of the processes through improved parameterization. This is required even more as the models are continuously being used at higher horizontal and vertical resolutions. The mojor challenges to the present day modelling community is,
a) The need of parameterization schemes which is scale aware,
b) Schemes that will be stochastic and also multiscale in nature
c) A unified scheme of physical processes.
The debate of cloud resolving approach or hybrid approach (with parameterized convection and grid scale cloud microphysics) and its impact on model fidelity will be emphasized.