Abstract

The effects of different model physics and different convective and boundary layer parameterization schemes are investigated using an 18-h nested-grid numerical simulation of the mesoscale convective systems (MCSs) that were responsible for the 19-20 July 1977 Johnstown flood. It is found that convective and resolable-scale diabatic process play crucial yet very different roles in the development and evolution of the MCSs. In particular, latent heat release resulted in development of strong vertical circulations, generation of an upper-level jet streak, formation of pronounced mesoβ-scale surface pressure perturbations, and rapid amplification of the traveling mesoα-scale wave that helped initiate the condensation processes. Resolvable-scale condensations appear to be directly responsible for the generation of a warm-core mesovortex and indirectly for a mesoscale convective complex (MCC). Without resolvable-scale heating, the model only reproduces the propagation of a squall line.

Incorporation of moist downdrafts also had a significant impact on the general evolution of the MCSs by producing important surface perturbations such as mesohighs and outflow boundaries. The role of moist down-drafts in the life cycle of the MCSs appears to be twofold. On one hand, the downdrafts vertically stabilized atmospheric columns and removed low-level moisture that otherwise would have been used for mesocyclogenesis and stratiform precipitation. On the other, the downdrafts horizontally destabilized the environment through the formation of horizontal temperature and pressure gradients. Specifically, it was found that the cold outflow boundaries over western and southern Pennsylvania helped the development and organization of continued deep convection during the nighttime hours. However, in central Pennsylvania, the warm-core mesovortex was significantly weakened when moist downdrafts were coupled with the updrafts in the convective parameterition scheme.

Inclusion of radiative heating in the surface energy budget tended to produce a conditionally unstable environment favorable for the development and maintenance of deep convection. In general, inclusion of the radiative heating at the surface improved the prediction of timing, frequency and location of convective precipitation. Omission of radiative heating has roughly the same “breaking” effect on the development of the mesovortex as the introduction of the moist downdrafts. It appears that the pronounced diurnal cycle of MCCs is directly related to the thermal cycle of the boundary layer.

Because of sharp and pronounced inhomogeneities in the horizontal moisture distribution, inclusion of virtual temperature instead of just temperature can considerably increase horizontal gradients of geopotential height. Without the virtual temperature effect in the ideal gas law, the model fails to reproduce the warm-core mesovortex and the MCC.

Use of a bulk boundary layer parameterization scheme appears to have a significant effect over mountainous regions. The scheme tends to overestimate the upward energy transport on the upslope side of a terrain feature and underestimate it on the downslope side.

In general, the results indicate that rigorous treatment of model physics is extremely important for simulating the mesoscale convective weather systems and precipitation associated with the Johnstown flood. The results also indicate that successful prediction of “convective” weather systems not only hinges upon the convective parameterization, but also upon the magnitude and distribution of the resolvable-scale latent heat release, and the concurrent development of the diurnal cycle of the boundary layer.