LES4CCFM: Using LES to characterize and parameterize the convective cloud field

点击翻译按钮获取中文摘要

The aim of this project is to use Large Eddy Simulations (LES) to characterize and quantify key elements of convection. The work will be performed within the framework of a new parameterization, the Convective Cloud Field Model (CCFM) coupled to the UK Met Office Unified Model (UM). Outcomes from this work will not only lead to a better CCFM but more crucially, will inform any new or existing convection parameterization for the UM. The main focus during the first three years, the exploratory phase, is the improved physical understanding of convection and its representation in CCFM. CCFM is a multi-plume convection scheme and described in Wagner and Graf (2010). Using an entraining parcel model the scheme calculates possible clouds of different initial radii for a given environment and large scale forcing. The distinctive feature of CCFM is the cloud spectrum calculation which determines the actual number of possible clouds. The spectrum calculation is described by a multivariate Lotka-Volterra system in which clouds compete for CAPE (convective available potential energy) through their cloud work function. From the individual clouds and their number, mass fluxes and thus convective heating and moistening can be derived. Initial tests of CCFM have been performed within the ECHAM climate model. In single column mode and in terms of precipitation timing and intensity ECHAM-CCFM performs significantly better than the standard ECHAM. Although no tuning has been applied yet ECHAM-CCFM improves many of the known precipitation biases in global simulations of the atmosphere with prescribed sea surface temperatures. As part of the proposed project we will implement CCFM as an additional option for the parameterization of convection. UM-CCFM will used to translate findings from the LES studies in other work packages and to evaluate the impact of changes in the convection parameterization on large scale dynamics. Since observations are often sparse and incomplete, LES driven by observations offer the best tool for the evaluation of CCFM. High-resolution modelling so far has focused on describing the cumulus ensemble. Despite progress a real breakthrough has not been possible since the cumulus ensemble is an average over very different entities that strongly interact and that is difficult parameterize directly. The proposed project LES4CCFM will overcome limitations of previous high resolution studies by explicitly investigating individual clouds and the resulting cloud spectrum separately. We will use higher spatial resolution (order decametres) than many previous studies to ensure that important parts of the mixing between clouds and their environment through entrainment and detrainment are explicitly resolved. LES studies will be performed in three consecutive work packages. First, we improve the representation of individual clouds in convection parameterizations by comparing LES output with prediction from the entraining parcel model in CCFM. LES to study entrainment, detrainment will start from a known, thus prescribed cloud base. Atmospheric profiles will be based on field campaigns and intensive field observations. In a second work package will focus on the convective cloud trigger by performing LES with a fully interactive boundary layer including surface fluxes, cloud microphysics and radiation. This work will replace with currently rather simple and ad hoc convective cloud trigger in CCFM with one that is more sophisticated and physical based. As part of a third work package we will characterize the convective cloud spectrum as it evolves over time and after it reaches equilibrium. We will evaluate the predator-prey assumption in the CCFM spectrum calculation. The outcome will not only be a new convection scheme within the UM with much broader physical basis but in addition will deliver new quantitative insight into convective processes that is crucial for any convection parameterization development.