How Zdrcal Works

The zdrcal data acquisition step resembles that of the Suncal utility, except that a PPI task consisting of one or more full rotations of the antenna is acquired, instead of automated sector scan around the anticipated sun position.

The zdrcal task is similar to a standard PPI task, except that all elevations are in the vertical position. The data outcome of the vertical scans are formatted as RAW product files for subsequent analysis.

The zdrcal data analysis step consists of retrieval of the RAW product containing the ZDR data fields subject to configurable IRIS/RDA quality thresholding, supplemented with standard moment data of reflectivity dBZ, Doppler velocity (V) and spectrum width (W). The data of co-polar correlation coefficient RhoHV completes the input to estimate ZDR offsets and associated uncertainties reported as dedicated result files.

Auxiliary Variables

Uses of the auxiliary variables dBZ, V, W, and RhoHV include:

  • zdrcal-specific quality criteria are applied gate-to-gate, to enhance the echoes of atmospheric scatterers in the ZDR data sample.

    The ground clutter residuals which may remain after Doppler filtering, and other spurious signals such as radio frequency interference, and macroscopic object echoes are diminished.

  • The expected variances of ZDR data values are computed as gate-to-gate sampling variations, by referencing the ZDR distribution as echoes from continuous precipitation in the sampling volume.1

ZDR Offsets Estimates

The ZDR offsets estimates are weighted medians of the qualified ZDR data points, checked for consistency with a maximum probability estimator of the same data set.

The check rejects estimates that appear disturbed by excessive non-Gaussian tails or other asymmetries biasing the median estimator from the most plausible ZDR offset value.

The weighting is based on expected gate-to-gate ZDR variances. Data points of high co-polar correlation and large spectral width are favored, which associate with narrowly distributed ZDR data. The weighting approach is statistically optimal in the sense that it maximizes the precision of the ZDR offset estimates, given variable input quality.

The ZDR offsets estimates are reported with estimates of uncertainty, which are derived from actual observed distributions of the ZDR data sample. Also, the input distributions are characterized in terms of expected and actually observed widths, allowing easy quality control. All estimates of uncertainty are expressed at 68 % confidence level ("one sigma" uncertainties, equivalent to root means square widths RMS of normally distributed data).

Post-processing

The zdrcal –process:file triggers a post-processing mode of zdrcal analysis using an existing zdrcal RAW product. The RAW product resides in the directory specified in the zdrcal.conf file.

The outcome of post processing is equivalent to the run-time result, pending the conditions that the RAW data are appended on the same ZDR history sample, and that the same zdrcal analysis settings are configured.

This is option is not used at most operational radars.

This command mode is useful for general quality control for example the effects of variations in quality criteria can be studied. The option allows for monitoring ZDRCAL calibration remotely for example at analysis centers of a radar network. Note that the RAW product does not need to originate from another Zdrcal utility run, but can be a standard IRIS RAW product, with requirement that the specified data dBZ, V, W, ZDR, RhoHV are available (and only those data fields).

1 The formalism in reference V.N. Bringi and V. Chandrasekar, 2001: Polarimetric Doppler Weather Radar, Principles and Applications, 2001. 636 pp. Cambridge University Press is applied.