Project No: 16301721

Title: The Representation of Turbulence and Convection in the Gray Zones of Orographic Precipitation

Principal Investigator: Prof. Xiaoming SHI


Topography modulates atmospheric flows in various ways and significantly influences the spatial patterns and intensity of the precipitation over mountain ranges, which is called orographic precipitation and essential for water resources in mountainous areas and downstream. Orographic extreme events, meanwhile, can cause disasters such as flooding and landslides to disrupt our society. The forecasting of orographic precipitation is thus an essential topic for numerical weather prediction (NWP). With advancing computing power, kilometer-scale resolutions have become increasingly popular in regional NWP, and in global simulations, it has also been experimented with recently. However, kilometer-scale grid spacing puts weather models into numerical simulation gray zones, the resolution range at which particular processes are neither entirely subgrid (i.e., unresolved) nor fully resolved. The application of kilometer-scale resolution for orographic precipitation, in particular, is within the gray zone for deep convection and possibly for complex terrains in certain locations. Partial resolving of flows in gray-zone simulations means that subgrid-scale (SGS) processes actively interact with the larger scales. Thus proper SGS parameterization is critical for accurately representing the cross-scale flow interactions in gray-zone simulations. However, traditional convection schemes designed for substantially finer or coarser grid spacings may use the assumptions that SGS turbulence is dissipative, isotropic, or horizontally homogeneous, none of which is appropriate for the gray-zone resolution. This predicament is the target of the proposed research. We will investigate how the effects of scale-aware convection schemes may differ from others in gray-zone simulations of the intense orographic rainfall embedded in a tropical cyclone environment. To make column-based convection schemes, which assumes horizontal homogeneity, more appropriate for gray zones, we will add a new component to model the horizontal mixing due to SGS turbulence. Idealized large-eddy simulations (LESs) and cloud-resolving simulations will be conducted to quantify the relation between the reliability of explicit convection, model resolution, and mountain geometry. We will also use machine learning to optimize the parameterization configuration according to specific weather patterns strategically. Overall, this project will advance our understanding of the gray zones of orographic precipitation and enhance our gray-zone prediction techniques.