Fine-scale numerical analysis of the sensitivity of the HyMeX IOP16a heavy precipitating event to the turbulent mixing-length parametrization.

This study investigates the issue of turbulence representation in the 'grey zone' at a sub-kilometric scale (500 m) and analyses the sensitivity of the simulated convective systems, their mesoscale environments and their triggering mechanisms. Special attention is paid to a case of heavy precipitation (IOP16a), which occurred on 26 October 2012 and was well documented during the HyMeX SOP1. To this end, sensitivity experiments were performed with the mesoscale non-hydrostatic French model Meso-NH to investigate the impact of different mixing-length formulations at a sub-kilometric scale. The simulations succeeded in representing the observed precipitation systems and the principal mechanisms of maintenance and formation such as low-level convergence and the low-level cold pool. Spatial precipitation distribution and system evolution are fairly well reproduced with a realistic surface rainfall pattern. However the simulated environment and convective processes are highly sensitive to the formulation of the mixing length. The developing and mature stage are similarly reproduced. Nevertheless the intensity of the low-level cold pool depends on the simulation. When the mixing length is larger, sub-grid turbulent kinetic energy (TKE) is larger, and the moisture advection and the accumulation of the hydrometeor species is reduced; in this case the system is less intense with less precipitation. Conversely, when the mixing length is weaker, sub-grid TKE is weaker and winds are increased to balance this effect. The simulated convective systems are more intense associated with larger moisture advection, higher hydrometeor contents and marked low-level cold pools. [ABSTRACT FROM AUTHOR]