WSP01796

• File Number: 8283.100
• Description: Colorado River Computer Models - Colorado River Simulation System
• State: CO
• Basin: Colorado Mainstem
• Water Division: 5
• Date: 7/18/1996
• Author: Unknown
• Title: Sufficiency of Colorado River Storage to Satisfy Consumptive Use - Draft
• Water Supply Pro - Doc Type: Report/Study

'.

OiHGuS

fOlmally adopted criteria. In the absence of such criteria, an analysis was performed to
examine the positive and negative effects of a 1996 surplus determination. To make the effects
more pronounced and enhance the analysis, the effects of arbitrarily determining a surplus for
an additional 1 0 years were also examined.

MODEL ASSUMPTIONS

To measure the effects of a surplus declaration, a NO SURPLUS case was assumed and ran.
The NO SURPLUS case assumed that there would be no surplus determination in 1996 or in
future years.

Starting Reservoir Contents
The CRSSez Model simulated operation of the Colorado River system starting from January 1,
1997 through January 1, 2050. Starting contents of system reservoirs were determined based
on the July 1 forecast of 1996 April-July Lake Powell inflow of7.45 MAF.
< .

Using the 1996 April-July inflow, the CRSSez Model assumed reservoir contents on January 1,
1997 to be:

Fontenelle
Flaming Gorge
Taylor Park
Blue Mesa
Navajo
Other Upper Basin Reservoirs
Powell
Mead

211 Thousand acre-feet (KAF)
3,289 KAF
75KAF
580 KAF
1,260 KAF
382 KAF
20,749KAF
21,496 KAF

These starting conditions assumed a surplus was issued for 1996. For NO SURPLUS runs the
1996 surplus was not assumed, and the starting contents for Lakes Mead and Powell were each
increased by 200 KAF to restore the effect of the release. Under the NO SURPLUS runs"the
starting contents for Lake Powell and Mead were assumed to be:

Powell
Mead

20949 KAF (3677.88 ft)
21696 KAF (1193.98 ft)

Hydrologic Assumptions
To conduct the analysis, the model simulated conditions for 54 years using traces developed
from the historic hydrologic data base. Using the 90 years of historical data, ninety traces of
54 years each were developed. Each trace would begin on a different year, and following the
1995 data, the data would be wrapped around to start again with 1906 data.

The weakness with using the historical data is that the historical pattern is repeated over and
over. Studies of older river systems do not find that the historical pattern repeats. Rather, a
stochastic method for developing possible future hydrologic traces is preferred. Using a
stochastic method the statistical characteristics of the historical data are preserved while year to