Abstract:
To better simulate the spatial characteristics and small-scale variations of wa ter substances in numerical models is of crucial importance for improving the nu merical weather prediction models, especially for precipitation forecast. With t he rapid development of workstation, the resolution of numerical model is incr easingly sophisticated, it is possible to calculate the cloud explicitly, so it bring forward a higher request for the computation of water substance advection, it must be high order accuracy, conservation, and shape preserving. The water substance field is positive definite scalar field, with the characteristics of l arge variations in space and time, strong gradients and not even continuous, the simulation of water substance is always a problem in numerical prediction.
The piece wise rational method is adopted as the scalar advection scheme in the GRAPES model, and this advection scheme is of highorder accuracy, conservati on, and shape preserving in solving the water substance advection. However, unph ysical phase change may occured over the computational areas of cloud boundary w hen coupling the advection scheme with precipitation processes. Such unreasonabl e features result from the semiLagrangian interpolation of cloud water around the cloud boundary and the consequent nonlinear interaction with physical proces ses. To mitigate this problem, a physical limitation method NLSL scheme in PRM s calar advection scheme in the GRAPES model was introduced, and tries to improve the simulation of cloud substances and precipitation forecast.Through one and tw o dimensional idealized experiments, the feasibility and effect of the physical limitation method in scalar advection scheme are verified. Furthermore, by imple menting this scheme into PRM advection scheme of highresolution GRAPES mesos cale model, both case study and continuous forecast experiment show that the simulated distribution of cloud water and thermal structures aro und the cloud boundary and within the cloud is improved to a certain extent. And , the forecast location of heavy rainfall and rain band exhibit more consistency with the observation.