WMOD00546

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

tried to obtain estimates of the average sampl ing error in determing areal
mean rainfall from precipitation gage measurements. The usual procedure is
to represent the sampl i ng error as the difference between the best est imate
of the true mean obtained from the maximum density of gages and sample means
calculated from successively reduced gage densities. Sampling errors calculatE!d
in this way are at best relative errors since the true mean is not known in
an absolute sense. The approa:ch used herein differs from those of previous
studies in that the isohyetal patterns of interest are not initially derived
from point measurements of pr,ecipitation but are theoretically generated by a
continuous spatial distribution function so that the true precipitation
characteristics are known quantities.

2. Approach

Storm characteristics such as type, structure, duration, and spel~d and
direction of movement affect the size, shape, and precipitation gradient of
the isohyetal pattern it produces. The distribution of surface rainfall will
vary greatly between storms of similar area and duration, particularly for
convective storms in which cells develop and coexist at various times and
locations during the storm's 'lifetime (Crane, 1979). The isohyetal pattern
of moving convective storms w'ill, therefore, frequently contain multiple
peaks and will rarely be regu'lar in shape. An example of a convE~ctive storm
isohyetal pattern is shown in Fig. 1.

The complexity and variability of convective storm isohyetal patterns make it
difficult to model the entire storm pattern in a general manner. However, we
can cons ider that the i sohyetal pattern of the storm is composed of smaller
isolated precipitation areas which are characterized by closed isohyets and

2