Josh Gebauer-April 20

Determining the Optimal Smoothing Parameter for Spatially Variable Advection Correction


April 20, 2018 - 2:00 pm


April 20, 2018 - 3:00 pm


120 David L. Boren Blvd., Room 5600, Norman, OK 73072   View map

Determining the Optimal Smoothing Parameter for Spatially Variable Advection Correction


Vertical velocity is the most difficult wind component to accurately retrieve from dual-Doppler analysis. The accuracy of retrieved vertical velocity from dual-Doppler techniques using only an anelastic mass conservation restraint is limited by absent low-level data and the lack of a proper boundary condition. Previous studies have shown that a variational dual-Doppler analysis that includes an anelastic vorticity constraint along with a mass conservation constraint provides more accurate vertical velocity retrievals, especially when low-level data is missing. The use of a vorticity constraint, however, requires an estimate of the local vorticity tendency. This local vorticity tendency cannot be calculated directly unless rapid (< 1 min) volume scans are available. Improvements in the estimation of the local vorticity tendency could further improve vertical velocity retrievals. One method for estimating the local vorticity tendency uses the advection velocities determined from a spatially variable advection correction technique. Spatially variable advection correction allows for the advection velocities to vary across the analysis domain, but the technique requires an arbitrary smoothing parameter. Currently, the optimal smoothing parameter for the advection correction technique is not known. Using RaXPol data from 24 May 2016, the optimal smoothing parameter was determined for varying volume scan times, analysis grid spacing, and analysis time steps. It was found that the optimal smoothing parameter increases with increasing grid size. The results also were sensitive to the analysis time step, particularly when the volume scan time increased. This sensitivity was a surprising result, but potentially could be caused by solution non-uniqueness.