Models of physical systems generate large volumes of results, usually in the form of numerical data that describe the solution of some differential equation by producing approximate results at discrete intervals in both space and time. The more refined the solution, the closer spaced these intervals will be, but one never generates a continuous solution of the form produced by analytic solutions, where the answer consists of an equation that can be evaluated at arbitrary points in space and time. Without this closed form solution, one must examine the results in their numerical form, since one does not have the form of the equation to examine. Lists of numbers are difficult to grasp, and one necessarily resorts to some graphical display of the solution in order to gain some insight into the nature of the solution. One-dimensional problems lend themselves to simple line graphs (X versus Y), which are familiar to most students. As one solves partial differential equations in two or more dimensions, the visualization of the solution becomes more difficult. In a two-dimensional domain, one needs a three-dimensional representation of the solution. Traditionally, two-dimensional contour or color-coded representations are used, but a three-dimensional topographical representation often produces a more realistic depiction of the solution. Topics in Scientific Computation (COS515), in both the finite-element and the computer modeling version, includes enough scientific visualization to allow the students to represent their solution graphically and get a feeling for the techniques and difficulties involved.