Estimating Flood Levels with Grapher

Dr. Rory Nathan (rory.nathan@unimelb.edu.au) is Associate Professor of Hydrology and Water Resources at the University of Melbourne, one of the leading universities in Australia. He is particularly interested in the stochastic modelling of natural systems and divides his time between consultation work on real-world problems and undertaking research, both while teaching environmental applied hydrology. Dr. Nathan has used Grapher for many years for exploratory data analysis and to display these technical results in his presentations and reports.

One such project was estimating flood levels using Monte Carlo analysis based on hydraulic modeling. Developing these estimates requires processing tens of thousands of simulations at a large number of locations. Dr. Nathan needed a way to quickly and heuristically verify the results at each location for internal consistency and to analyze the influence of outliers. Additionally, the generated reports would then be used in formal reporting, so the quality of the images was very important.

Dr. Nathan made heavy use of Grapher’s scripting abilities to automate processing these numerous reports. With some basic scripting knowledge and help from the Golden Software’s user forums, he created a script that read in data from multiple ASCII files, created the plots, adjusted the site-labeling and y-axis limits on each individual plot, and then exported the plots as graphical images. 100MB of data took approximately five minutes for Grapher to process. Thereafter, Dr. Nathan was able to verify the results at a later point and was able to use the plots directly in his final report.

When asked why he’s continued to use Grapher all these years, Dr. Nathan replied, “I have always valued the precise control over the display characteristics of the [Grapher] plot. The program allows me to tell the (graphical) story exactly how I intend