Supplementary Modeling Study of a Tropical Convective Band

View More View Less
  • 1 Department of Atmospheric Sciences, University f Illinois, Urbana, Illinois
© Get Permissions Rent on DeepDyve
Restricted access

Abstract

A simulation is made for the observed evolution of a mesoscale tropical convective band using a nonhydrostatic cloud model and compared with an earlier Ogura and Kao simulation where a hydrostatic model was used with the Arakawa-Schubert cumulus parameterization scheme. Both simulations are made in two dimensions, and identical large-scale forcing that is time-independent but spatially varies is imposed in the form of low-level convergence. No environmental wind is included in the models, aside from the wind representing the imposed large-scale low-level convergence. In Ogura-Kao's simulation, the parameterized convective activity occurs all the time during the simulation period in or near the area of the maximum large-scale forcing. In sharp contrast, our model-resolved convective clouds start developing in the area of the maximum large-scale forcing, but then propagate away like a multicellular (but slowly moving) squall-line from that area, as a result of interactions between the cloud-induced cold outflow and the low-level wind shear associated with the imposed large-scale forcing. Evaporation from the ocean surface is found to be not fast enough so that the area of the maximum large-scale forcing is left as a rain-void area during the 24 h simulation period, except for the initial convectively active period. The result indicates the correct precipitation simulation obtained in Ogura-Kao may be regarded as fortuitous.

Abstract

A simulation is made for the observed evolution of a mesoscale tropical convective band using a nonhydrostatic cloud model and compared with an earlier Ogura and Kao simulation where a hydrostatic model was used with the Arakawa-Schubert cumulus parameterization scheme. Both simulations are made in two dimensions, and identical large-scale forcing that is time-independent but spatially varies is imposed in the form of low-level convergence. No environmental wind is included in the models, aside from the wind representing the imposed large-scale low-level convergence. In Ogura-Kao's simulation, the parameterized convective activity occurs all the time during the simulation period in or near the area of the maximum large-scale forcing. In sharp contrast, our model-resolved convective clouds start developing in the area of the maximum large-scale forcing, but then propagate away like a multicellular (but slowly moving) squall-line from that area, as a result of interactions between the cloud-induced cold outflow and the low-level wind shear associated with the imposed large-scale forcing. Evaporation from the ocean surface is found to be not fast enough so that the area of the maximum large-scale forcing is left as a rain-void area during the 24 h simulation period, except for the initial convectively active period. The result indicates the correct precipitation simulation obtained in Ogura-Kao may be regarded as fortuitous.

Save