WMOD00563

• Title: Snow Accumulation Algorithm for the WSR-80D Radar: Final Report
• Date: 7/1/1998
• Weather Modification - Doc Type: Report

300
KCLE .
KEN X .
250 KFTG*
0 KGJX ..
. KMPX 0
N
II 200
0
-+-
Q)
CD 150
~
0
-
o 100
.s::.
a.
<(
50 ~
.. ~
*
0
0 25 50 75 100 125 150
Range Km

Figure 5.-Values of a for individual gages, all for 13 = 2.0, show a linear decrease with range for five
different radars. Linear regression lines are shown for each radar's data set.

vary among the different data sets. The greatest slope is with the Cleveland lake effect storm data,
known to be shallow with steep vertical profiles of snowfall. The least (and markedly different) slope is
for Albany where large synoptic storms predominated, producing relatively heavier snowfall rates. The
shallow Denver upslope storms and weaker Minnesota synoptic storms have similar intermediate slopes.

The only serious outlier from the 5 linear relationships of figure 5 is gage # 2 of the KMPX line near
60 km range. That gage had a lowest maximum hourly accumulation of the 5 Minnesota gages over the
entire 1996-97 winter at only 0.060 inch (next lowest was 0.085 inch). This difference suggests gage # 2
was subject to more wind undercatch than the other gages. Use of equation (3) would show that a lesser
gage catch would result in a greater a value for a fixed ~ (see table 1).

Plots similar to those of figure 5 were made (not shown) with the same ordinate but with the abscissa
being the vertical distance of the center of the lowest available beam above each surface observing site.
These plots also showed a high degree of linear association albeit with somewhat more scatter.

6.7 Recommended values for a and 13

In view of the results presented thus far, the recommended Ze-S relations for the five regions represented
by the above radars are given in table 5. The a values have been rounded off to the nearest ten from
table 4 since there is no practical significance in listing more precise values, unlikely to have the implied
accuracy.

28