제목: Development of a Stochastic Convection Paramterization and its application to Climate Modeling​​​​​​

While there are many challenges associated with developing convection parameterization, one major issue that has recently gained attention is stochastic convection parameterization. In this study, we investigate the physical processes that generate convective cloud variabilities and develop a stochastic parameterization that simulates the mean and variance of convective tendencies based on the unified convection scheme (UNICON). By extending UNICON, we develop a stochastic UNICON with convective updraft plumes at the surface randomly sampled from the correlated multivariate Gaussian distribution for updraft vertical velocity and thermodynamic scalars. The updraft plume radius at the surface follows a power-law distribution with a specified scale break radius. In addition to the stochastic initialization at the near-surface, a stochastic mixing model with a machine learning technique is proposed. Stochastic UNICON is tested using a single column model and a global climate model, and the simulation results indicates that stochastic UNICON improves the simulation of mean states and intraseasonal variabilities.​