Two novel, general purpose algorithms for unfolding Doppler velocities

Reliable unfolding of aliased Doppler velocities in cloud and weather radar volumes is requisite before this data is assimilated into numerical weather prediction (NWP) and other high-resolution models. In addition, the quality of the corrected radial velocities can have considerable effects on retrievals including multi-Doppler derived winds which make use of this dealiased data. Numerous algorithms for dealiasing velocities have been proposed, but most of these algorithms are tuned for the setup of a particular radar. Few, if any, of these algorithms are capable of correcting velocities from multiple scan strategies and a variety of operational wavelengths.

In this talk, two novel dealiasing algorithms will be introduced which are designed to be general purpose in nature and can be applied to PPI and RHI scans as well as data collected from vertically pointing instruments. These algorithms are applicable to all radar wavelengths and Nyquist velocities. Both of these algorithms are implemented and available for immediate use in the open source Python ARM Radar Toolkit (Py-ART).

The first algorithm makes use of a multi-dimensional phase unfolding method originally designed to analyze optical fringe-patterns. This involved a gate-by-gate unfolding of the field following a path based on the local reliability of the gate. The second algorithm unfolds velocities in a similar manner to the procedure used when manually dealiasing radar volumes. Namely, regions of similar velocities are identified and then unfolded against each other by modeling the system as a dynamic weighted graph. This method results in a global minimization of all velocity differences between detected regions. This talk will focus on the technical details of these algorithms and provide examples of their use to correct velocity aliasing in radars operated by the Department of Energy's Atmospheric Radiation Measurement (ARM) program.