SACRCOR

This dataset contains moments from the Scanning ARM Cloud Radars (SACRs) which have been filtered and corrected in order to provide a product for non-radar meteorologists which contains fewer measurement artifacts and effects from signal propagation.

Corrections have been applied to selected moments. The 'reflectivity_corrected' variable has been corrected for gaseous attenuation effects. Dealiasing of Doppler velocities is performed using two algorithms, with the results stored in the variables 'mean_doppler_velocity_corrected' and 'mean_doppler_velocity_corrected2.'

This product removes returns from non-atmospheric source, such as insects, ground clutter, and second trip echoes, leaving only meaningful returns from clouds and precipitation.  Users should note that the filter mask in this product is not perfect.  A small amount of clutter, insects and second trip echo artifacts may still be present in the filtered data, especially near the radar where classification is particularly challenging.  Additionally, due to the instrument configuration, no filtering of insects was performed on data from the X-band SACR.  Improvements to these algorithms are planned as further testing is performed and additional data becomes available. The filter applied to the reflectivity is also applied to both corrected mean Doppler velocity variables.  This filter is appropriate for filtering other radar moments contained in this product but the task of extracting this filter and applying it to these additional moments is left to the user.  By not filtering all radar moments it is possible to examine both the filtered and unfiltered radar moments using a single data product.

Finally, the locations of each sweep in the radar volume are detected by examining the elevation and azimuth angles. These locations are stored in the same variables as in the a1 level files, simply prefixed by the string 'a1_.'

As mentioned earlier, two independent algorithms were used to dealias the Doppler velocities.  This duplication is intended to solicit feedback from users on the performance and reliability of these two algorithms to direct future development.  Users should note that the algorithm used to perform the dealiasing corresponding to the 'mean_doppler_velocity_corrected' variable performs unfolding ray-by-ray and provides a flag indicating whether the algorithm succeeded in dealiasing each individual ray.   These flags are stored in the 'velocity_unfolding_flag' variable.   Rays for which this variable indicates that dealiasing was unsuccessful, should not relied upon.  The second dealiasing algorithm unfolds velocities by finding regions of similar velocities in two dimensions and unfolds these regions  in order to minimize the difference between regions. Results are stored in the field 'mean_doppler_velocity_corrected2.'  Users should note that this algorithm does not use data from external sources such as soundings for comparison and therefore may report velocities which globally differ from the true velocities in the atmosphere by multiples of the Nyquist interval.  Additionally, in some cases the algorithm does not handle the periodic nature of PPI scans properly which may lead to subpar unfolding across  0 degrees azimuth.

Input datastreams for the SACRCOR product are the corresponding sacr.a1 datastream, the ceil.b1 datastream for insect detection and the mergesonde1mace.c1 datastream for dealiasing. Many of the algorithms used in processing are described in the 2014 article by Kollias et al. Feedback on this product from users is greatly appreciated and will help drive future development.