Bias in topographic thresholds for gully heads

Inter-event time series of seismic activities of Sumatra region with magnitude (m _b  ≥ 4.8) and period spanning over 1973–2012 has been analyzed using the techniques of nonlinear dynamics. The earthquake data were collected from the USGS Web catalog. The rank order statistics of inter-event time series reveal mixed response with distinct breaks in slope suggesting that earthquake dynamics in this region are unstable, but “self-organized.” Comparison of return maps of the data with random, stochastic, and chaotic time records shows quasi-deterministic behavior. We assessed the dimensionality of earthquake-generating mechanism using a nonlinear predictor technique on two-dimensional phase portrait constructed by recurrence time series. The nonlinear forecasting analysis suggests that the earthquake processes in the Sumatra region evolve on a non-random low-dimensional chaotic plane. Further, second-order Kolmogorov entropy “K2” analysis revealed a coherent time structure, indicating quasi-deterministic dynamical pattern. Our result is consistent with “self-organized” processes determined by the internal dynamics, where impulsively derived interdependencies cascade through the tectonic stress generated by plate movement. However, we do not preclude the triggering role of other external processes on the coupled underlying system.