An overview of advanced modeling through simulations
Release 9.3 Last modified January 5, 2009	Print all topics in : "Advanced modeling through simulations"

About this topic

Until now, geoprocessing and using ModelBuilder have been presented as ways to automate workflows that you encounter in your day-to-day work. This topic is a bit different—it's about using geoprocessing and ModelBuilder to simulate actual events in the landscape such as a the spreading of a fire or an oil spill. Simulations such as these allow you to explore the physical processes that drive the event such as wind, temperature, slope, vegetation, and ocean currents. The ultimate goal of simulation is to predict events so to better plan and mitigate their impact.

Most of the techniques used here, such as the use of lists, series, and model iteration, are documented elsewhere. It's how these techniques are used in simulation that is of interest. However, for certain simulations, you will be introduced to new capabilities such as calculating random numbers, assigning random values to each cell in a raster, and randomly placing a specified number of points.

About simulation modeling

The types of advanced models covered in this section include process models and process models with stochastic (random) events along with performing error and sensitivity analysis.

A process model models a physical event. A simulation model of an oil spill is an example of a dynamic process model. Some phenomena are not understood well, so they appear to be random such as the case with fire spotting. Fire spotting occurs every so often when a small piece of a fire breaks from the main body and jumps to another location. A simulation model of a fire spotting situation is considered a dynamic process model with a stochastic event.

Uncertainty is inevitable in all models. Uncertainty derived from error in the input data or the model parameters and its effects on the output can be explored through error analysis. And through sensitivity analysis, you can explore the influence of each parameter on the output.

For additional information on process models, process models with stochastic events, error analysis, and sensitivity analysis, see the following links:

General concepts of modeling through simulations
Creating process models
Creating models to capture dynamic stochastic events
Performing error analysis
Performing sensitivity analysis
References

Please visit the Feedback page to comment or give suggestions on ArcGIS Desktop Help.
Copyright © Environmental Systems Research Institute, Inc.