We demonstrate that some chaotic parameter subsets of a class of spatiotemporal chaotic systems modeled by globally coupled maps are riddled. That is, for every point in the chaotic parameter subset, there are parameter values arbitrarily nearby that lead to nonchaotic attractors. A consequence is an extremely sensitive parameter dependence characterized by a significant probability of error in numerical computation of asymptotic attractors, regardless of the precision with which parameters are specified.
- fat fractals
- large numbers