18th Conference on Weather and Forecasting, 14th Conference on Numerical Weather Prediction, and Ninth Conference on Mesoscale Processes

Monday, 30 July 2001
Response of Convection to Dry Layers: Simulation and Parameterization Tests
James A. Ridout, NRL, Monterey, CA
Poster PDF (120.6 kB)
Some effects of dry layers aloft on deep convection have been investigated using quasi-cloud resolving model forecasts for the region near the edge of a dry tongue in the western tropical Pacific. The forecasts were carried out using a triple-nested version of COAMPS, the Naval Research Laboratory's Coupled Ocean/Atmosphere Mesoscale Prediction System. The inner mesh in the COAMPS forecasts has a horizontal dimension of 3km, allowing for a somewhat crude resolution of deep convective circulations. Forcing data taken from the COAMPS 3km mesh grid were saved from the forecasts for use in convective parameterization tests in both single-column model (SCM) and semi-prognostic model (SPM) experiments. During a 24 hour forecast, a number of isolated deep convective cells break out in the 3km mesh grid. The deepest of these penetrate to about the 150mb level. To investigate the impact of entrainment of dry upper level air, the forecast is repeated using a moisture sounding from a TOGA-COARE dry tongue episode as boundary conditions for the inner mesh above a threshold pressure level. Several forecasts are carried out, with this threshold level varying between 800mb and 400mb. The entrainment of the drier air results in more than a 40 percent reduction in rainfall when the drier air is restricted to above 600mb. The peak cloud top heights are not appreciably affected. Results for these cases from SCM and SPM tests using several different convective parameterizations will be presented illustrating the effectiveness of the parameterizations in incorporating the effects of entrainment of dry upper level air. The significance of the horizontal resolution of the forcing data used for the column tests will be demonstrated.

Supplementary URL: