WSPC07048

• Title: MODSIM vs STATEMOD - Recommendation for CRDSS Water Resource Planning Model
• Description: Colorado River Basin - Colorado River Computer Models-Colorado River Decision Support System
• Decision Support - Doc Type: Report
• Date: 2/16/1995
• DSS Category: Surface Water
• DSS: All
• Basin: Statewide

O()005~

modeling concern, For example, an infeasible solution in the MODSIM
prototype took about a month to resolve partly because the identified
location of the infeasible solution (a reservoir) had no relation to the actual
problem (incorrect return flow rate on a demand). In another example, the
DNR team, even with three or four explanations by the consultant team,
were unable to understand why modeled return flows were not a direct
function of diversions.

STATE MOD uses an easy to understand solution methodology (addition
and subtraction). During development and further refinement of this
model in the past several months the DNR team has found it easy to
identify source of problems and correct those problems. We believe the
understandability of this model is reflected in the fact that the
implementation of a ST A TEMOD prototype in the Gunnison River Basin
took about two man months of effort compared with over 25 man months
(excluding time of W.W, Wheeler) of effort to implement a MODSIM
prototype,

Because of the ease of use and understandability of ST A TEMOD, the
consultant team should find that upcoming development of applications
for the remaining west slope basins should be more efficient with
STATEMOD than with MODSIM even though the consultant team at this
time is more familiar with MODSIM than STATEMOD.

The DNR team believes that future maintenance of the water resource
planning model is dependent on the understandability of the model
concepts. The DNR team believes that it could maintain the ST A TEMOD
model and applications in the future but questions its ability to maintain
the MODSIM model or applications.

EXECUTION SPEED:

MODSIM
t/t/

STATEMOD
t/t/t/

BASIS:

Execution time is always important on a multiuser, multitasking system,
The execution time of the water resource planning model is especially
important since the Gunnison Basin prototype represents only one basin of
the six to be modeled,

The execution speed of an application should only be compared when one
has two models applied to a similar problem. An "apples to apples"

5