WMOD00546

• Project Name: High Plains Cooperative Program
• Title: On the Sampling Variance of Raingage Networks
• Weather Modification - Doc Type: Report

Table 5. - Percentage contribution of total sample size to detect
a 25-percent change in mean Montana storm preci pi tat i on due to
sampling variance as a function of gage density (based on both
the median area of all raincells, AI, and the median area of
the primary raincells, A2)

GPR

Gage density
(km2/gage)

Percentage contribution
by sampling variance

Al

A2

100
50
10
5
2
1
0.5

0.9
1.8
9.0
18.0
45.0
90.0
180.0

1..6
3..2
16..0
32..0
80..0
160..0
320..0

0.05
0.05
0.5
1.9
9.9
26.9
50.7

Vi ewed in thi s manner it is cl ear that the natural vari abil ity of convective

precipitation is mainly responsible for the large sample size requirements in
evaluating the hypothetical precipitation augmentation experiment. The
sampling variance component is \~esponsible for less than 10 percent of the
total sample size requirement with a gage density of at least two per raincell
(an average of four per storm) and less than 27 percent with a gagE~ density
of only one per raincell (an average of two per storm). Based on the observed
raincell areas and their contribution to the storm's total rainfalll, this
corresponds to moderately dense raingage networks of 45 to 80 km 2 per gage
and 90 to 160 km 2 per gage, respectively. In view of the large natural

variability, there does not appE!ar to be any advantage in using hi~,her
density gage networks. The main thrust should, for the present, bE! in the
direction of reducing the natured variability by physically meanin~lful
stratHications according to prE!dictor variables and covariates. It should,

however, be noted that as the efforts to reduce the natural vari abil ity
succeed, the importance of the sampling variance will increase and higher

22