We describe a linear modeling technique that derives functions (parameters) of the predictor variables in a tabular form using a Fourier transformation technique. The result is a complex multiple regression model that is actually a series of models independently describing each parameter. This technique (1) provides an approximation to nonlinear modeling that is not restricted to a small number of variables, (2) does not require precisely formulated theoretical functions for each variable, and (3) avoids many of the restricting assumptions normally required for nonlinear modeling. The estimate of the total systematic information (variance) in the system and the amount of information predicted by the model can be used as a measure of the efficacy of the model. A production form of the model is being completed for use on the Missouri River mainstem reservoir system for predicting the effects of reservoir management practices on the indigenous fish populations.