A novel framework for evaluating and improving parameterized subtropical marine boundary layer cloudiness

Abstract

A single-column model (SCM) is used to simulate a variety of environmental conditions between LosAngeles, California, and Hawaii in order to identify physical elements of parameterizations that are requiredto reproduce the observed behavior of marine boundary layer (MBL) cloudiness. The SCM is composed ofthe JPL eddy-diffusivity/mass-flux (EDMF) mixing formulation and the RRTMG radiation model. Modelforcings are provided by the Modern-Era Retrospective Analysis for Research and Applications, version 2(MERRA2). Simulated low cloud cover (LCC), rain rate, albedo, and liquid water path are compared tocollocated pixel-level observations from A-Train satellites. This framework ensures that the JPL EDMF isable to simulate a continuum of real-world conditions. First, the JPL EDMF is shown to reproduce theobserved mean LCC as a function of lower-tropospheric stability. Joint probability distributions of lowertropospheric cloud fraction, height, and lower-tropospheric stability (LTS) show that the JPL EDMFimproves upon its MERRA2 input but struggles to match the frequency of observed intermediate-range LCC. We then illustrate the physical roles of plume lateral entrainment and eddy-diffusivity mixing length inproducing a realistic behavior of LCC as a function of LTS. In low-LTS conditions, LCC is mostly sensitiveto the ability of convection to mix moist air out of the MBL. In high-LTS conditions, LCC is also sensitiveto the turbulent mixing of free-tropospheric air into the MBL. In the intermediate LTS regime typical ofstratocumulus-cumulus transition there is proportional sensitivity to both mixing mechanisms, emphasizingthe utility of a combined eddy-diffusivity/mass-flux approach for representing mixing processes.