Comparison between a Three-Dimensional Simulation and Doppler Radar Data of a Tropical Squall Line: Transports of Mass, Momentum, Heat, and Moisture

Abstract

Results from a detailed three-dimensional cloud model are extensively compared with Doppler radar data in the case of a fast-moving tropical squall line, observed during the COPT81 experiment. The comparisons use a two-dimensional statistical analysis, in which we consider each four-dimensional field as the sum of a two-dimensional averaged vertical field and a four-dimensional residual field, representative of along-line and temporal fluctuations. These fluctuations are analyzed by computing the time average of spatial variances and the temporal variance of spatial averages. The results indicate good agreement in both quality and quantity, particularly for the vertical velocity and line-normal wind. The consequences of this agreement are two-fold. First, a physical and comprehensive model of the convective part of a tropical squall line emerges that is coherent with both observations and simulations. This conceptual model emphasizes the importance of along-line fluctuations. Second, it is possible to use the simulation results to compute thermodynamic budgets and vertical transports of mass and momentum. Both observations and simulations confirm that, for the convective part, the vertical transport of line-normal momentum is predominant relative to the along-line one. The line-normal transport cannot be either neglected or simply explained in terms of diffusion. The mean scale lifting and rear-to-front mean circulation give the mean structure of the momentum flux vertical profile. However, the transport due to convective eddies is important and serves to extend vertically the total transport. Except in the first kilometer, the effect of total advection is a loss of momentum at midlevels and a gain aloft. The pressure gradient across the convective part acts in opposition to the advection but is insufficient to neutralize it and to reach, at this scale, a steady state in the squall line's moving frame. For both observations and simulation, that results in a weak evolution of the convective part's structure. The latent heating provides a good approximation of the apparent heating source; nevertheless, computation of this latent heating should include rain evaporation. The computation of the apparent moisture sink necessitates that of its convective eddy transport. The transport due to convective eddies decouples the vertical distributions of heat and moisture. Normalization by the rainfall rate allowed us to compare successfully the simulated apparent heating source and moisture sink with previous diagnostic studies with larger time and space scales. We thus confirmed the double-peak structure of apparent humidity sink Q2 that is often seen in tropical budget composite studies. However, we propose a different interpretation through the vertical transport of Q2 by convective eddies.