WSP05071

• File Number: 8146.400
• Description: Pueblo Dam - Reports
• State: CO
• Basin: Arkansas
• Water Division: 2
• Date: 10/1/1994
• Author: US Geological Survey
• Title: Use of Frequency Analysis and the Extended Streamflow Prediction Procedure to Estimate Evacuation Dates for the Joint-Use Pool of Pueblo Reservoir Colorado
• Water Supply Pro - Doc Type: Report/Study

3022

uation date, and (2) summing the calculated release
discharge volumes from the evacuation date bacle in
time until the cumulated contents was about equal to
the capacity of the JUP. The curves shown in figure 21,
which are intended to provide an approximation of the
maximum allowable JUP contents, are based on
0.01 EP daily mean discharges in Fountain Creek and
the St. Charles River.

FORECASTING INFLOW VOLUME TO
PUEBLO RESERVOIR USING THE
EXTENDED STREAMFLOW PREDICTION
PROCEDURE

Historically, discharge-volume forecasts usually
have been made using regression techniques that corre-
late recorded discharge volumes to measured water
equivalent at snow courses in and adjacent to a basin.
These forecasts usually are made for April I through
September 30. Although an April-September forecast
possibly could have been incorporated into the devel-
opment of figure 17, the many factors that can affect
discharge volume for this longer period probably
would have increased the complexity of applying the
study results.
To provide the capability to forecast April 15-
May 14 inflow volume to Pueblo Reservoir, the opera-
tional capability of the NWSRFS model was applied to
the Arkansas River basin upstream from Pueblo Reser-
voir. The reservoir inflow forecasts are made using the
ESP procedure, which is a component of the NWSRFS
model. The National Weather Service routinely uses
the NWSRFS model to provide short- and long-tenn
forecasts for a variety of uses by National, State, and
local agencies, and by the general public. Therefore,
forecasts of the April15-May 14 inllow volume to
Pueblo Reservoir would be readily available for any
year in which the PRIUP study results would be
applied. The following sections of this report present
brief descriptions of (I) the NWSRFS model, (2) the
ESP procedure, and (3) examples of ESP discharge vol-
ume forecasts. Detailed descriptions of these topics
can be found in the cited references.

National Weather Service River Forecast
System Model

Initial development of the NWSRFS model is
described by the National Weather Service (1972);
however, continued development and improvement of
the model have resulted in a dynamic, multi-volume
users manual that is continually revised and updated by
the National Weather Service. The NWSRFS model

has three major components, the calibration system
component, the operational forecast system compo-
nent, and the ESP component. The following descrip-
tions are derived from Day (1985) and Anderson
(1986).

The calibration system primarily is a collection
of programs that are mathematical representations of
various components of the hydrologic cycle. The two
primary components that were calibrated for the
PRlUP study are a snow-accumulation and -ablation
model (Anderson, 1973) and a soil-moisture account-
ing model (Burnash and others, 1973). To ensure that
streamflow simulated by the NWSRFS model com-
pares closely to recorded streamflow, values for the
model parameters (the coefficients of the mathematical
representations) were optimized by using trial-and-
error or computer techniques, or a combination of both
(Brazil and Hudlow, 198 I).

The calibration system also includes preproces-
sor programs that manipulate data used in the calibra-
tion program and compute mean areal precipitation and
mean areal temperature time series data for input to the
calibration programs. A 39-year (water years 1949
through 1987) data base of discharge, precipitation,
temperature, diversion, and reservoir-contents data was
compiled for use in calibrating the NWSRFS model
and in implementing the ESP procedure for the PRlUP
study. Although discharge data are available for longer
periods (table I), the number and the locations of pre-
cipitation and temperature stations prior to about 1948
are inadequate for model calibration.

The operational forecast system then uses the
calibrated model components with real-time hydro me-
teorological data to provide forecasts of discharge. The
operational forecast system contains (I) a data-entry
component, (2) a preprocessor component, and (3) the
forecast component. The data-entry component assim-
ilates real-time data from various sources and then for-
mats the data for input to the preprocessor component.
The preprocessor component estimates missing data
and also calculates the mean areal precipitation and
mean areal temperature time-series data for input to the
forecast component. The forecast component then per-
fonns the hydrologic and hydraulic simulations needed
to make the forecast.

Use of the operational forecast system to make a
discharge forecast requires estimation of precipitation
and temperature data for the forecast period. Because
these data cannot be estimated reliably for more than a
few days into the future, the operational forecast sys-
tem cannot be used to make forecasts for more than a
few days into the future (Day, 1985, p. 158). Develop-
ment of the ESP procedure (Twedt and others, 1977;

FORECASTING INFLOW VOLUME TO PUEBLO RESERVOIR USING THE EXTENDED STREAMFLOW PREDICTION 41
PROCEDURE