Determination of Bias in Zdr

S-Pol MAP Italy, 1999

Importance of Bias Determination

Note that high accuracy in Zdr is desired when Zdr is applied to precipitation accumulation estimates. For precipitation estimates, a change of 0.3 dB in Zdr will result in a 17% change in estimated accumulation.

Since S-Pol is a research radar system, particular attention is given to ensuring the accuracy of all parameters. Every effort is made to minimize the error in Zdr, with the goal of obtaining accuracies of .2 dB, or better.

Technique

To estimate Zdr bias, we require a well-characterized, non-biased, system-independent data set for analysis. Such a data set can be obtained from viewing rainfall at vertical incidence.

For purposes of Zdr bias estimation, there is no consistent axial asymmetry in raindrop shape when viewed from directly below. In the case of a strongly sheared environment, there might arguably be preferential orientation of a collection of hydrometeors. Fortunately, preferential orientation effects can be reliably removed by rotating the vertically pointing antenna in azimuth and averaging results over several rotations. It may also be surmised that snow crystals can serve for bias determination, however, experience with frozen hydrometeors shows somewhat greater variablility in results, and there was one particular problem during MAP that reduces the utility of frozen hydrometeors for bias determination.

Practical Considerations

• light precipitation at the radar
• precipitation-sized hydrometeors extend vertically for several kilometers
• radar dish not too wet from previous precipitation
• signal power return is not too great

• data collection period of 2 minutes or greater
• antenna is slowly rotated while looking vertically
• clutter filter is OFF (error on this during MAP)

Every attempt is made to process appropriate cases in a very consistent way. Vertical-pointing scans are transferred to a data processing system. A script is used to drive the SOLO analysis package, performing parameter thresholding and general data selection, through production of a histogram of Zdr values for qualifying data. The only operator interaction is in the definition of a broad boundary to time-window the data and to eliminate obvious artifacts.

Table 1 summarizes the automatic thresholding criteria applied to vertical pointing data.

TABLE 1. Vertical Pointing for Zdr Bias Determination: threshold criteria

Description	Relation	Reason
eliminate data when received power is too high	DM > -45.0 dBm	avoid regions of non-linear receiver response
eliminate data in regions of weak reflectivity	DZ < ~0 dBZ (flexible criteria)	avoid refractive index scattering or any other low-reflectivity oddities
eliminate bright band and regions of wet hydrometeors	LDR > -13.0 dB	Zdr can have a very wide distribution in these regions
avoid regions close to radar	range < 1.2 km	eliminates regions of TR tube recovery, and selects data in far-field, only
remove data above the atmosphere	range > 14.0 km	also eliminates inadvertant inclusion of test pulse

Analysis

For MAP, twenty-one separate events of vertical pointing were logged. Of these, about one-third were considered to be ultimately unsuitable. Within the 21 events, subsets were deliniated and processed separately. These subsets provide an indicator of the short-term consistency of the procedure and results. All events are listed in Table 2. Each independent event is separated by a horizontal line; within each event, any subsets are shown on their own line. For determination of gross conclusions, subsets are first combined to provide a single set of event values. Within the table, links are provided to either an image of the data (time section of received power, DM, and thresholded Zdr, TZDR), or the histogram produced in the analysis. Futhermore, events and their subsets have been subjected to multiple analysis in an attempt to separate the characteristics based on hydrometeor type (ice or liquid; upper or lower layer).

The Results Section follows Table 2.

TABLE 2. Determined values of Zdr Bias, by event.

Date	UTC	Default mean sdev #pts	Ice/Upper mean sdev #pts	Liquid/Lower mean sdev #pts	Zdr Test Pulse,dBm	Comment
990904	1039 1042	0.12 0.37 40859 0.09 0.35 54194	0.07 0.36 28147 0.08 0.37 31465	0.25 0.14 3069 0.21 0.24 3878	unkwn
990906	1215				-.77	not usable
990913	0800					not usable
990913	1310					not usable
990916	0737 0740 0743 0746 0755			0.03 0.12 1782 0.02 0.14 5940 0.02 0.15 10738 -0.01 0.16 9130 -0.04 0.21 3834	-1.00	assume liquid
990917	2111 2114 2117 2120	-0.01 0.18 73582 0.00 0.19 105701 0.01 0.20 103268 0.01 0.21 96329	-0.04 0.15 44473 -0.03 0.18 50629 -0.02 0.20 44647 -0.02 0.21 30431	0.08 0.19 17648 0.05 0.19 28834 0.06 0.20 30631 0.06 0.20 32570	-.92
990918	0719 0722		-0.01 0.29 37035 -0.01 0.28 25682		-.92
990925	1806	0.10 0.26 19751	0.05 0.30 9729	0.15 0.20 10250	-.90
990930	0815 0819	0.19 0.12 3579 0.21 0.11 3983			-.88
991003	0135 0139 0143	0.18 0.28 21151 0.14 0.29 12330 0.18 0.21 27754	0.15 0.30 14494       0.13 0.23 16332	0.26 0.20 4601       0.28 0.16 9155	-.86	weak ice/liq distinction
991003	2104 2108	0.19 0.35 26358 0.21 0.33 11848	0.11 0.37 10194 0.11 0.35 4042	0.29 0.24 5225 0.27 0.22 6024	-.88	weak ice/liq distinction
Feedhorn changed out after ops of 991003
991013	1219 1223 1237	0.16 0.43 22119	0.01 0.48 18860 0.03 0.43 21237 0.08 0.35 6249	0.19 0.45 15957	-1.09	v little liq v little liq poor distinc.
991017	1120 1123 1136	0.19 0.27 5037 0.31 0.24 293     14			-1.11 -0.96 -1.05	OK very poor not usable
991018	1448 1451		0.02 0.45 30707 0.02 0.45 32026		-1.04	probably ice probably ice
991020	1757	0.19 0.44 33570	0.16 0.43 27149	0.37 0.45 4081	-.83
991020	1854 1857	0.27 0.40 16935 0.29 0.43 15219	0.21 0.41 10909 0.22 0.43 9134	0.39 0.24 2229 0.47 0.25 1768	-.83	liquid distribution is poor
991020	2055	0.20 0.44 30468	0.15 0.40 21515	0.36 0.27 2748	-.85
991021	1656	0.19 0.43 36142	0.15 0.46 28100	0.32 0.19 6747	-.87
991023	1209	0.48 0.25 271			-.88	a poor case
991103	1010 1015 1020 1025	0.33 0.42 42369 0.33 0.42 39223 0.33 0.42 41039 0.32 0.41 38531	0.27 0.47 30038 0.27 0.46 27086 0.28 0.46 28309 0.27 0.44 26564	0.48 0.17 7628 0.48 0.18 7511 0.48 0.19 7931 0.45 0.25 10819	-.78
991104	0801 0805	0.34 0.39 17717 0.35 0.39 17062	0.27 0.43 10978 0.25 0.43 9213	0.46 0.17 4791 0.48 0.25 6753	-0.77

Results: Zdr Bias for MAP

• An error was made in leaving the clutter filter "ON" during MAP Zdr bias determinations. Work needs to be done to estimate typical bias for vertical pointing in rain due to the clutter filter, and to understand why such a rain/snow bias should exist in the first place. It is also possible that the perceived rain/snow difference has nothing to do with the clutter, but is an effect of differential recovery of the S-Pol T/R tubes.

• S-Pol uses two T/R (transmit-receive) tubes, one for each transmitted polarization. These tubes contain a gas that ionizes during the strong transmit pulse, protecting the receiver components from damage. Recovery from ionization takes a finite amount of time, with most recovery (to within 3 dB of non-ionized tube) occuring within 1 km of the radar. For MAP, the two tubes recovered at slightly different rates, with the resulting differential recovery likely causing the difference observed between the ice and liquid hydrometeor regimes. Generally, the T/R tubes may be considered fully recovered after a range of 6 km, but most vertical pointing observations were performed for particles closer to the radar than this.

• Within a single event, Zdr bias determinations are very reproducible (mean of about +/- .03 dB)
• The Zdr bias varies with date and time.
• Zdr bias over a half-day period varies very little, and may be considered constant within that timeframe (variability of +/- .02 dB; see supporting information; likely exception on 991017)
• There is a difference in the bias in Zdr between particle types (ice or liquid); this is likely due to range effects (related to T/R tube recovery).
• Zdr bias characteristics change somewhat with the change of antenna feedhorn.
• The majority of Zdr bias change is closely correlated to the change in Zdr test pulse value (particularly true from 991013, onward). Inter-event reproducibility is on the order of +/- .05 dB, after corrections are made using the test pulse.
• Final, absolute accuracy of corrected Zdr is estimated on the order of +/- 0.20 dB, with some small, additional uncertainty introduced by miscellaneous effects.

• Zdr bias drift was not evident in previous two S-Pol deployments; cause is being investigated.
• A table will be developed detailing the applied Zdr bias correction vs date/time.
• There is some small hint in prior experiments of different Zdr bias for different particle types or radar range, but the effect is much smaller than that seen for MAP (rain/snow or near/far difference of < .05 dB).