WSPC00727

• File Number: 8283.200
• Description: Colorado River Basin-Colorado River Computer Models-Colorado River Decision Support System-Ray
• State: CO
• Water Division: 5
• Date: 11/24/1993
• Title: AOP-Memorandum to File-AOP Model vs CRDSS
• Water Supply Pro - Doc Type: Report/Study

r

~'

)001207

I

CRDSS MEMORANDUM

TO:
FROM:
OATE:
SUBJECT:

CROSS File
Ross Bethel
November 24, 1993
AOP Model vs. CROSS

J

On November 4th, 1993, Randy Seaholm, Ray Alvarado and I met with
Terry Fulp to see the USBR's implementation of the AOP model on a
workstation. The USBR spent approximately $ 175,000 with CADSWES
to develop a GUI for the FORTRAN AOP model (on a SUN workstation).
This implementation was performed several years ago and is still
used on a frequent basis at the Upper and Lower Region level.

Terry downloaded the model and some sample data sets from Brad
Vickers at the Upper Region. Even though Terry had not used the
AOP model in several years, it appeared to be a relatively easy
exercise to demonstrate the model. He showed us how to pull in
data sets, edit those data sets, specify how to develop graphics of
interest. Terry demonstrated how to present three graphics on the
screen at a time, and be able to interactively alter one of the
graphics of input data (i.e. reservoir releases), operate the model
and view the changes on the other graphics. This appeared to be a
very effective tool for looking at various operating strategies.

We discussed the graphic functions on the SUN workstation that may
be difficult to convert to a SGI platform. The graphics use the
GKS functions which is somewhat archaic (before X-windows and
Motif) . Terry indicated the ease of a conversion would be
dependent on the availability of GKS functions for the SGI. In
subsequent discussions with state Malers of RTI, Steve indicated
that he had downloaded the GKS functions from MIT and that they
should work fine with a SGI system. This would facilitate the
conversion of the AOP graphics functions to SGI.

We discussed how best to incorporate the AOP into the CROSS. Terry
again indicated that the USBR is planning on using the new Big
River model (to be developed) for the AOP but that AOP development
will only occur after full development of the Big River Model and
also development of a daily operating model version of the Big
River Model. The development of a new AOP model will definitely
not occur in the first year, possibly in the second year of USBR
activities but possibly later. Given the demonstration of the AOP
workstation version, steve Malers perception of GKS function
availability for the SGI, and the indefinite USBR schedule, it is
my feeling at this time that it would be productive to port the AOP
workstation version to the SGI for use in CROSS. The graphics are
quite effective and could quickly lend to the perception of
significant progress on CROSS. This is to be explored in more
detail during Phase 1.