The Building Blocks of Vertical Velocity in Deep Convection over Darwin, Australia and Their Evolution with the Convective Population

Poster PDF

Description

The co-location of a 50-MHz profiler, a 920-MHz profiler, and the C-Band polarimetric radar (CPOL) radar in Darwin, Australia presents an unique opportunity to investigate how vertical velocity in deep tropical convection is related to the physical characteristics of the convection. Williams [2012] retrieves unbiased vertical velocity data by using the 920-MHz profiler to filter out hydrometeor fall velocities from the 50-MHz data. The horizontal scanning strategy employed by the CPOL radar enables physical characteristics of convection passing over the profilers to be determined. These characteristics include the depth, size, and intensity of the precipitation. This study uses data from the 2005-06 and 2006-07 wet seasons. First, an empirical orthogonal function (EOF) analysis is conducted on the vertical velocity data to identify the dominant structures in the vertical velocity profiles. Over 80% of the variability in the vertical velocity data can be accounted for by the three leading EOFs. Considering all combinations of the positive and negative principle components associated with each EOF allows us to identify eight canonical vertical velocity profiles. Four profiles are dominated by strong upward motion aloft. Two profiles are characterized by weak downward motion near the surface and moderate upward motion aloft. Two profiles have very weak vertical motion through the depth of the convection. Finally, we investigate how convection evolves through these eight vertical velocity profiles as it passes over the profiler. Two dominant vertical velocity evolution pathways are identified. Vertical velocities increasing in magnitude and depth with time characterize one pathway. The second pathway is characterized by weakening of the two profiles with moderate upward motion aloft and weak downward motion near the surface. Analysis of the CPOL data reveals that these vertical velocity pathways are associated with changes in the physical characteristics of the convection, including variations in the depth and slope of the echo top heights, the hydrometeor types, and the profiler’s proximity to the edge of the precipitation feature. This work charts a path forward to ultimately retrieve vertical velocities across the entire CPOL radar domain based on their observed physical characteristics.