WMOD00564

• Title: Snow Accumulation Algorithm for the WSR-80D Radar: Supplemental Report
• Date: 11/1/1999
• Weather Modification - Doc Type: Report

most of their precipitation growth in the layers close to the ground. To avoid ground clutter, higher tilt
radar data had to be used. However, that overshoots the growth zone, resulting in the need for smaller
alpha values to calculate the correct surface precipitation.

All these straight lines present a mathematical problem. A range correction would be the ratio of a
standard alpha, such as 150, and the alpha values along the line. The corrections would go infinite
(inverse of zero) where the lines intersect the range axis and then negative at farther ranges. Therefore,
straight lines are inappropriate for range correction, even though the lines are derived from real, precision
surface measurements. For comparison, the NIDS version of the SAA had a quadratic relationship for a
range correction factor, as described below in Section 5.2. That relation is converted to an equivalent
function of alpha varying with range and is shown as the thick curved line in figure 3. Alpha is 150 until
36 km. Thereafter, alpha changes to a parabolic curve resulting in a correction factor of about 3 (alpha of
50) at 230 km. The parabolic relationship avoids infinite corrections but is a distortion of the actual
relationship.

Table 2.-Data used in the derivation of a range correction for the
Minneapolis NEXRAD radar
Tilt Meters R,km Ratio Inv.
0.50 378 35.0 1.000 1.000
1.45 958 73.4 0.890 1.124
2.40 1,538 103.7 0.780 1.282
3.35 2,117 129.5 0.670 1 .493
4.30 2,696 152.4 0.560 1.786

col1 = NEXRAD antenna tilt angle - degrees - VCP21
col2= center of beam above radar at 35 km range (meters)
col 3= range,km to 0.5 deg beam center for altitude of col 2
col 4= ratio of S to 35 km, 0.5 deg. tilt value; linear to 0.56 at 4.3 deg.
col 5= correction factor (reciprocal of col 4)

5.2 Range Correction from a Seasonal Averaged Vertical Profile of S

The range correction scheme previously developed for nine Minnesota snow storms of the 1996-97
winter, described by Super (1998), was revisited. All the radar data were used in new calculations with a
somewhat different Ze -S relationship discussed by Super and Holroyd (1998). The results were quite
similar, with the median vertical profile of Ze suggesting about a 20 percent decrease in Ze per kilometer
of height above the ground. Alternatively, a seasonal average or median gradient can be derived for S
rather than Ze.

8