Pavlos Kollias — Stony Brook University
Scott Matthew Collis — Argonne National Laboratory
Kirk North — McGill University
Jonathan Helmus — Argonne National Laboratory
Category
ARM Infrastructure
Description
Figure showing unfolding of a single sweep from an ARM XSAPR radar located at the SGP. Panels show the raw, aliased Doppler velocities, the unfolded velocities resulting from running the FourDD algorithm, the multi-dimentional phase unwrapping algorithm and the region based unfolding algorithm.
Radial Doppler velocities measured by ARM’s radars are limited in range due to the nature of the measurement. When the velocity of the atmosphere is outside of this range, the value measured by the radar is folded or aliased into this limited range. Various routines and algorithms have been designed to unfold or dealias the measured velocities to determine the true atmospheric velocities. When these algorithms are applied to Doppler velocity data from ARM’s radars, the results are often imperfect and contain obvious artifacts.
In this poster, we present initial results from two novel dealiasing algorithms being developed for use in future ARM Value Added Products. The first uses a multi-dimensional phase unfolding algorithm originally designed to analyze optical fringe-patterns. The second unfolds velocities in a similar manner to the procedure used when hand 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.
Examples will be presented comparing these two new algorithms with an existing published algorithm on data from ARM’s scanning precipitation radars (XSAPR and CSAPR) and scanning cloud radars (SACR) as well as on data from non-ARM radars. All three algorithms are available for testing in the Python ARM Radar Toolkit (Py-ART)
