The authors provide an overview of agent-based simulation (ABS) for environmental modelers, with the purpose of linking user requirements to currently available techniques. Although the paper contributes positively to the classification of a restricted number of techniques for a limited group of users, it is suggested that a broader-based taxonomy of ABS would have been even more constructive.

A recent SciFinder search for ABS models recovered more than 500 published reports. It seems that a larger sample could be used in some formal classification procedures, to provide a more general classification, which would also be useful to the more imaginative environmental modelers. The need for more effective agent-based capabilities is evident from the proliferation of Internet sites with large amounts of data and information stored in heterogeneous sources. An ABS taxonomy that is more inclusive of other fields seems likely to benefit all potential users, including environmental modelers.