21st Conf. on Severe Local Storms and 19th Conf. on Weather Analysis and Forecasting/15th Conf. on Numerical Weather Prediction

Tuesday, 13 August 2002: 4:30 PM
The impact of parameterized shallow convection on pre-deep-convective sounding structures in the Eta model
John S. Kain, CIMMS/Univ. of Oklahoma and NOAA/NSSL, Norman, OK; and M. E. Baldwin, S. J. Weiss, and M. P. Kay
Poster PDF (228.5 kB)
Shallow convection is parameterized in numerical models to simulate the vertical mixing and transport effects of non-precipitating cumulus clouds. Since a shallow convection parameterization (SCP) does not produce precipitation, it can be quite difficult to determine where it has been active during a model forecast. Yet, a SCP can have a dramatic (and sometimes unrealistic) impact on sounding structures and can strongly influence a forecasters' opinion regarding the probability of both the initiation and potential severity of deep convective. Thus, forecasters can benefit from being able to recognize when an SCP has been active and what its potential impact on a model sounding has been.

Operational forecasters at the SPC have become relatively sophisticated in their ability to recognize shallow convective activity in the Eta model. Their skill in model-sounding interpretation has proven to be quite valuable in recent months because the SCP in the Eta has apparently played a significant role in the overprediction of CAPE and the erroneous prediction of other thermodynamic fields in some convective environments. The Betts-Miller-Janjic convective scheme in this model aggressively activates shallow convection and appears to spread its influence over an unrealistically deep layer. By warming the bottom half of this layer and cooling the top half, the scheme destabilizes the atmosphere and appears to produce exaggerated values of CAPE. This model behavior will be documented through the presentation of several case studies and comparison with observed soundings and an experimental version of the Eta model containing the Kain-Fritsch convective parameterization. The implications for the prediction of convective weather will be discussed.

This behavior has been noted in past years, but appears to be occurring more frequently of late. We have developed a hypothesis that this problem has been exacerbated by the recent resolution increase in the Eta model. Data to either support or refute this hypothesis will be collected to presented at the conference.

Supplementary URL: