WMOD00308 ‎(2)‎

• Title: Operation and Evaluation Report of the Santa Barbara County Cloud Seeding Program, 1977-78 Winter Season
• Date: 8/1/1978
• State: CA
• Weather Modification - Doc Type: Report

4-17

~

and in the unseeded control areas of Santa Barbara and San Luis Obispo
Counties. Stations with lU1:i.mpaired flO\oJs \vi thout upstream diversions
were found to be scarce but three stations were found in the control
area and nine gaging stations were located in arKl near the target.
Station names and numbers are given in Table 7. Figure 8 provides the
locations of these stations.

Historical streamflow data (in acre-feet) were tabulated for these
stations for the period of January and February. In order to exclude any
effects that might have occurred during the research project (1967-1974)
the seven years of data for these months were elirninated from the sample.
This meant that the period of record for most of the stations was quite
short.

~~ltip1e linear regression equations were co~~puted for target and
control area runoff. The 1978 streamflow was then compared to the pre-
dicted runoff expected from the regression equations. While in general
the multiple correlations were good the results were not (See Table 8).
This is apparently due to the fact that most of these streams have little
or no flow most of the time. Very few cases of high flO\.,r~ such as was
observed in 1978, are contained in the historical sample. Thus the regres-
sion is based upon isolated events and tends to either overpredict or
underpredict the streamflow in the extreme years S1l1ch as 1978. To illus-
trate this effect, scatter diagrams of the target stations versus a single
control station were prepared. These diagrams did show that almost all
of the historical data points were grouped together near the low-flow end
of the line with only one or two isolated points at the other end. It is
interesting to note that two stream gages, namely numbers 1395 and 1400,
which are very close to each other (See Figure 8) indicate ratios (based
on calculated versus observed flow) of .44 and 1.53, respectively. Since
they are both part of the Sisquoc River drainage it does not seem reason-
able that tile ratios of calculated runoff should differ by nearly 100%: