Using Lagrangian parcel trajectories in a large-eddy simulation of deep convection to evaluate a downdraft parameterization

Abstract:

Precipitation-driven convective downdrafts are represented with varying degrees of complexity in most cumulus parameterizations. They are associated with convective cold pools, and reduce the stabilizing effects on the environment of updraft mass fluxes.

We are using a large-eddy simulation (giga LES) of deep tropical oceanic convection to investigate the key assumptions used in the downdraft parameterization of the Integrated Forecasting System (IFS) model developed at ECMWF. These assumptions include the downdraft origination level, the downdraft mass flux at the origination level, and the profiles of downdraft properties below the origination level. The IFS assumes that evaporation of convective rain maintains a saturated descent.

We collected 146,800,640 Lagrangian parcel trajectories (LPTs) at 30-s intervals from the giga LES. LPTs avoid problems that can arise from analyzing Eulerian downdraft statistics. The Lagrangian perspective is especially valuable for understanding the interactions of entrainment, detrainment, turbulent mixing, and microphysics in downdrafts.

The graphic shows the average relative humidity of downdraft trajectories that end below 500 m and that originate within height intervals of 1-2 km, 2-3 km, and 3-4 km.