Laserfiche WebLink
<br />003114 <br /> <br />Page 13 <br /> <br />III. <br /> <br />METHOD <br /> <br />A. Conceptual Framework <br /> <br />A formal model which can be implemented on a computer is a virtual <br /> <br />necessity in a study such as this one because it permits rapid, repeated, and <br /> <br />inexpensive exploration of a complex set of assumptions about the values of <br /> <br />uncertain parameters. TWo main options are available. The first is to use a <br /> <br />generally-structured model which has been developed previously for its power in <br /> <br />solving a prescribed class of problems. Many such general models are available. <br /> <br />The choice of which one to select depends upon the correspondence between the <br /> <br />characteristics of the problem under investigation and the structure of the <br /> <br />model. After it has been chosen. such a model must be specified with empirical <br /> <br />data derived from the situation which is under study. <br /> <br />The second approach is to create an entirely new model by specifying not <br /> <br />only the empirical data or facts of the situation under study but also the <br /> <br />mathematical representations of the relationships between those "facts." A <br /> <br />model developed under this approach, if it is a good one, will explain the <br /> <br />phenomena under study better than a generalized mathematical model, but it will <br /> <br />normally be far less tractable to use. Thus, general models are often preferred <br /> <br />for exploratory or reconnaissance investigations while situation-specific models <br /> <br />are developed subsequently when the problem is better defined. <br /> <br />, <br /> <br />Considerations such as these led to the selection of a generalized network <br /> <br />optimization model in this first phase exploration of possible uses of enhanced <br /> <br />streamflows in the Colorado River system. <br />