Collaborative Construction and Visualization of Scientific Models

Principal Investigators:

Prof. J. R. Miller, EECS
Prof. J. J. Feddema, Geography
Prof. T. A. Slocum, Geography
Prof. D. C. Cliburn, Hanover College

The interface between science and decision-making is one of the greatest challenges in information technology research today. Political decision makers need sound scientific data when making decisions that may have, for example, significant impacts on the environment. Sophisticated computer models can be used to establish scenarios and investigate options, but these simulations are beyond the understanding of most legislators. Worse yet, the models as well as the data on which they operate have error and uncertainty. Therefore visualization of scientific models must necessarily include a visualization of some measure of the uncertainty in the results so that decision makers will be able to understand not only our predictions, but also how confident we are in them.

The collaborative decision support environment we have created allows political decision makers, members of their staffs, and domain experts to collaborate in an attempt to develop good public policy. We currently focus on same time, same place collaboration, but two other ongoing efforts are moving this capability into the same time, different place arena. Different time, different place collaboration is next.

Example 1

Using RGB colors to indicate uncertainty in model results. Black (i.e., absence of color) implies high confidence (i.e., absence of uncertainty). Mixtures of red, green, and blue indicate relative amounts of uncertainty due to three independent sources.

Example 2

The raw results of a prediction of future drying due to global climate change Illustrating confidence in the results of the model by varying the transparency of the surface according to the amount of uncertainty present at the location.