Using Agent-Based Modeling to Investigate the Emergent Properties of Storm Chasers

College of Letters and Science / Atmospheric Sciences / Mathematical Sciences


Seimon et al. (2016) posited, following a study of the infamous El Reno, OK tornadic event, that scientific teams interested in coordinated storm chases for data collection could take advantage of the availability of non-professional storm chasers. They suggested this based on detailed analysis of the location and movement of the numerous storm chasers that pursued this storm, finding that the collective emergent behavior of the group was ideal from the standpoint of data collection. Emergent properties of individual agents are well known from such natural phenomena as flocking birds, and is an active area of my current research. Working with the student, we will build an agent-based model of a portion of Oklahoma (including highways, county roads, and dirt roads) and track the properties of clusters of storm chase vehicles as a function of assumptions concerning their decision processes and storm characteristics (this is a concept similar to one I am pursuing for considering the quality of storm evacuations in advance of hurricanes). The results of this study could be of considerable interest to the meteorological community as the coordination of scientific storm chase teams has been an ongoing problem in field studies.

Tasks and Responsibilities

The student will assist me on several fronts: (1) collection of basic data for use in the model, such as road systems; (2) model-building (in the course of which he will be trained by me in Fortran coding); (3) model experiments. Throughout this work, we will be consulting with the authors of Seimon et al. (2016), who are already aware of this project. This will give the student an opportunity to interface with active members of the meteorological research community.

Desired Qualifications

None listed