Modelling Global Climate Variables with Cellular Automata Networks

Robinson, David (2015) Modelling Global Climate Variables with Cellular Automata Networks. Undergraduate thesis, under the direction of Dr. Dawn E. Wilkins from Computer and Information Science, The University of Mississippi.


Download (5MB) | Preview


Cellular automata are discrete models that can be used to simulate many physical systems. Cellular automata have been used to model gas diffusion, different types of chemical reactions, population growth, and land use change over time. Recent research into cellular automata networks has shown that if sparse long range connections are added to a cellular automata, then it will exhibit properties of complex networks. Furthermore, research into modeling climate systems has shown that modeling the global climate as a complex network can be used to predict individual climate variables. In this work we attempt to connect these ideas by simulating global climate variables, from the National Center for Environmental Prediction / National Center for Atmospheric Research Reanalysis 1 Dataset, as a cellular automata model and as a cellular automata network model. In our experiments we use neural networks as the cellular automata transition functions, using both single and multi-variable data. The results of our work suggest that cellular automata networks are better at modeling climate variables than standard cellular automata and that cellular automata based modeling is a viable approach to modeling climate data.

Item Type: Thesis (Undergraduate)
Creators: Robinson, David
Student's Degree Program(s): B.S. Computer Science
Thesis Advisor: Dr. Dawn E. Wilkins
Thesis Advisor's Department: Computer and Information Science
Institution: The University of Mississippi
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Depositing User: Mr Caleb Robinson
Date Deposited: 06 May 2015 19:20
Last Modified: 06 May 2015 19:20

Actions (login required)

View Item View Item