Morgan Schneider - April 9

 Convective Meteorology (Mesoscale Dynamics)   Quantifying and Mitigating Debris-Induced Bias in Doppler Radar Measurements of Tornadic Winds Morgan Schneider Friday, April 9th 3:00pm   Join Google Meet: https://meet.google.com/ksh-txvg-kni Lofting and centrifuging of tornadic debris are known to cause bias in Doppler radar measurements of tornado wind speeds.  Debris presence within

Start

April 9, 2021 - 3:00 pm

End

April 9, 2021 - 4:00 pm

 Convective Meteorology (Mesoscale Dynamics)

 

Quantifying and Mitigating Debris-Induced Bias in Doppler Radar Measurements of Tornadic Winds

Morgan Schneider

Friday, April 9th

3:00pm

 

Join Google Meet:

https://meet.google.com/ksh-txvg-kni

Lofting and centrifuging of tornadic debris are known to cause bias in Doppler radar measurements of tornado wind speeds.  Debris presence within a resolution volume is consistently associated with anomalous positive radial divergence as well as underestimation of both azimuthal and vertical velocities.  We chose to use a simulation-based framework to study these errors.  In the first part of this work, polarimetric radar time-series simulations from SimRadar are populated with debris of varying type, size, and concentration to independently examine how debris characteristics contribute to error in Doppler velocity and retrieved wind fields.  To estimate bias magnitude, velocities calculated from these simulations are compared one-to-one with velocities obtained from a “ground truth” no-debris simulation.  We then correlate the spatial distributions of velocity bias magnitude with those of relevant moment and polarimetric variables (reflectivity factor, differential reflect
ivity, and co-polar cross-correlation coefficient).  These relationships can be used to qualitatively estimate the likelihood that a given radar volume suffers substantial velocity bias.

The second part of this work presents an attempt to identify and subsequently remove debris contamination within Doppler spectra, thereby correcting the associated velocity errors.  This method incorporates dual-polarization spectral density (DPSD) estimates to examine the spectral polarimetric characteristics of scatterers within a resolution volume.  Debris presence is identified in the spectral domain using simplified HCA fuzzy logic.  Debris-dominated spectral components are then filtered from the original signal, and the Doppler velocity is recalculated using the new reconstructed signal.  Using the same techniques as the error quantification in part I, the filtered velocities are compared to the original debris-less simulation to determine if the DPSD filtering method is effective at mitigating debris-induced Doppler velocity bias.  These algorithms show promising potential for improving radar velocity estimates, and will be applied to observational datasets from the Rapid-scan
ning X-band Polarimetric (RaXPol) mobile radar in the future.