WMOD00527

• Title: Quantitative Precipitation Forecasting for Improving Reservoir Operations
• Date: 4/1/1995
• Weather Modification - Doc Type: Report

00 50 100 150 200 25Cb
I.{) L.()
N N
POD = 11/14 = .79
CJ) FAR = 10/21 = .48
u..
(.) 0 0
- 0 r = .87 0
0 N Std. Error = 20.7 k-cfs N
en
-0 Y = .87X - 5.9
c
ro *
en
:::J *
0 0 0
.c
l- I.{) * L.()
---- ..-
~
ro
Q)
a...
~
0 0 0
0;:: * 0
c 0
..- * ..-
E
0
en
0 * *
u.. *
-0 0 0
Q)
c: I.{) * * L.()
* *
Q)
en *.
.0
0 *
0 0
0 50 100 150 200 250
Predicted Folsom Inflow Peak (Thousands of C.F.S.)
Figure 5. - Scatterplot of observed inflow peaks to Folsom Reservoir with predicted inflow peaks based upon orographic
precipitation model estimates (C.F.S = fe/s).

as 205,000 ft3/s. The POD (probability of detection) of peaks of 50,000 ft3/s or more was 0.79;
i.e., 11 of the 14 observed occurrences were predicted. The FAR (false alarm rate) for
predicting these large flow values was 0.48; i.e., of 21 cases predicted, 10 were false alarms.
Both the correlation coefficient and the POD are encouragingly high. The FAR is acceptable,
though higher than desired. Also encouraging is the regression line slope and the small
intercept value.

An equivalent regression analysis using predicted peaks computed with observed
precipitation as HED7l input showed a correlation coefficient of 0.92, a line slope of 0.97, and
an intercept of -1600 fe/so The corresponding standard error of estimate was 16,500 fe/s, the
POD was 1.00, and the FAR was 0.26 (fig. 6). Each of these numbers indicate better
agreement between predicted and observed inflow peaks when using observed rather then
model-computed precipitation as HED71 model input, as might be expected.

To more directly compare the effects of using orographic model precipitation to using observed
precipitation, a regression analysis was performed on the predicted inflow peaks computed
by these two methods. This analysis removes any HED71 characteristics that might be
affecting the results of comparing to observed inflow. An encouraging correlation coefficient

15