One of the challenges in operational ensemble model prediction is to display results in an operationally useful form. Large amounts of model ensemble output and information need to be made more useful to forecasters without cumbersome or overwhelming displays. The graphical package, Vis5d and/or D3D provide a solution to the data compaction problem allowing the forecaster to view ensemble information efficiently. These graphical display packages allow animated 3-dimensional renderings and access to data in the 3 spatial coordinates, time and parameter space. Qualitatively, 3-dimensional iso-surface rendering serves as a tool for determining regions and times where the ensemble model spread or variability is minimized. When displayed in 3-dimensions, renderings of iso-surfaces show the number of ensemble components predicting a particular weather event, and trherefore, act as a proxy for determining the probability of forecaster selected meteorological events.

An ensemble of NCEP and national center operational model forecasts is presented in a Vis-5D display format for the parameters of height, temperature, relative humidity, wind, mean sea level pressure and precipitation. The 3-dimensional animated display system is configured such that the vertical coordinate is the ith ensemble forecast. This display provides a solution to many challenging aspects that face forecasters using ensemble predictions. The goal is the compaction of the large amount of ensemble model output, as well as a method to display standard maps and products for the individual forecasts without becoming cumbersome or overwhelming. A measure for determining regions where the ensemble model spread or variability is minimized showing areas where the models have greater confidence and affording a proxy for determining the probability of user selected meteorological events.

The Vis5D or D3D visualization display systems can be used to view data sets as iso-surfaces, and as standard contour plots. This is used to view the 3-dimensional animated rendering of the volume and depth of the convergence patterns along cyclones and fronts which we hypothesize can indicate the location and intensity of severe weather. 3-dimensional views of the size and shape of the convergence patterns appear correlated to severe weather as measured by observed winds and radar intensity. To test the usefulness of this hypotheses quantitatively, model predictions of convergence patterns form the NCEP global model will be analyzed for volume and vertical extent. The 3-dimensional animated convergence renderings as shown by Vis5D will be compared with composite radar observations.