By subjecting agents to a simplified environmental conditions, it is possible to clearly identify the behaviours that do not align with expectations.
An example of this form of verification can be found in the Rainfall model. While the model is designed to operate using a GIS raster input, which defines the behaviour of raindrops in relation to their flow and accumulation within the model environment.
We also allow abstract spaces. These include flat plane, a cone shape, and a simple hill as shown in the figure below
These "simplified" environments allow the movement of agents to be easily seen, and performance judged accordingly.
Experiment with the Rainfall Model. If you change the landscape (MapType) to Flat, Cone, and Hill, does the water behave as expected?
Click on the image below to see a YouTube movie of the verification options which act as a verification exercise about the basic functioning of the model (i.e. did we build the model right?). Here we have three different map types. The first being flat, the second being a cone and the third being a hill. The idea with these map types is to ensure the basic functioning of the model is correct. :
