WSP07957

• File Number: 8276.100
• Description: Grand Valley Unit-Colorado River Basin Salinity Control Project
• State: CO
• Basin: Colorado Mainstem
• Water Division: 5
• Date: 1/21/1983
• Title: Grand Valley Salt Pickup Calculations - Draft Report
• Water Supply Pro - Doc Type: Report/Study

(,.,
00
1.""
f;:;:;

~-;

.

.

.

v,

Sensitivity Analysis - A sensitivity analysis was performed on

the program using the data for the Cameo Station. The sensitivity

analysis results for Cameo should be applicable to all the river

stations because they all have about the same extent of data,

The parameters varied were MAXIHI, MAXILO, PCT1, PCT2, and the

coefficients in the TDS-EC linear regression equation. MAXIHI,

MAXILO, and PCT2 control the selection of the data availability case

(see Figure 13) and were discussed previously. PCTl is a tolerance

level specifying the maximum allowed percentage difference between

each grab sample 's TDS and the TDS calculated from the long-term

TDS-EC regression equation using the grab sample's EC. If PCTl is

exceeded, that particular grab sample will be ignored as inaccurate.

The best values of these parameters for Grand Valley river stations

are thought to be: MAXIHI = 65 days, MAXILO = 35 days, PCTl = 0.25,

and PCT2 = .25, resulting in a mean annual loading at Cameo of

1,425,855 tons/year. Table 40 displays the effect on the mean annual

salt load of independently varying these parameters. It can be seen

that the salt loading is not at all sensitive to thes,e parameters

which affect the choice of methods used to estimate l,)ading.

In addition to the aforementioned parameters, the sen,sitivity of salt

loading due to varying the TDS-EC linear regression c')efficients was

investigated. To determine this sensitivity, the slo:?" of the TDS-EC

regression line was changed plus and minus 10 and 15 percent and the

y-intercept adjusted such that the line still passed through the point

A-ll