Improving modeled momentum flux in the atmospheric boundary layer

Tropical cyclones (TCs), shallow cumuli, and low-level jets (LLJs) are all important phenomena in the climate system, but have been historically difficult to represent in climate models. For instance, even at higher resolution, simulated TCs often exhibit an incorrect relationship between minimum pressure and surface wind speed. Simulated shallow cumuli often exhibit a local maximum (“jet”) in the wind profile that too broad and diffuse. Simulated LLJs often suffer from a weak diurnal cycle of surface winds. All three climate model deficiencies may be related in part to inadequate parameterization of subgrid momentum fluxes in the atmospheric boundary layer.
The parameterizations of momentum flux in current-generation climate models are crude. Often momentum parameterization suites consists of downgradient diffusion plus a separate cumulus momentum transport scheme. However, the presence of a near-surface jet in the wind profile can sometimes lead to upgradient momentum flux at the top of the jet maximum, even when deep convection is not present. Furthermore, the need to model the complex diurnal evolution of winds in LLJs is difficult when the task of simulating momentum is divided between separate parameterizations.
This project proposes to parameterize momentum transport by prognosing subgrid momentum fluxes directly. This approach is quite different from conventional approaches, but it adheres more closely the governing equations, and hence is more flexible and general. For instance, it is capable of predicting upgradient momentum fluxes.
In this project, the process of momentum transport will be examined using a comprehensive hierarchy of observations and models. Based on these studies and improved understanding, prognostic equations for momentum fluxes will be refined and tested. The equations will be implemented into two leading climate models, one from GFDL and the other from NCAR. The resulting simulations will be evaluated against observations.