An empirical time series (1941–2007) of advance and retreat data from the coast of Guyana is modeled with statistical time-series techniques. Subseries of 5-y periods are fitted to modified Box-Jenkins space–time models. Second-order spatial-cyclic autoregressive models, associated with cyclical advance and retreat patterns, fit the data for five different subseries. First-order autoregressive models are also suitable to describe the data from five other subseries, thereby suggesting a long-memory response in the coastal system. Three of the subseries are fitted to space–time autoregressive moving-average models, thereby indicating the presence of random shocks (i.e., random events) in the coastal system. The various models are indicative of cyclical, long-memory, and short-memory processes operating in the coastal system. These processes can be associated with mudshoal propagation and stabilization and with temporal stochastic processes that force the coast to advance or retreat in different locations.
How to translate text using browser tools
1 November 2011
Time-Series Modeling of Data on Coastline Advance and Retreat
Sajid Rashid Ahmad,
V. Chris Lakhan
ACCESS THE FULL ARTICLE
Journal of Coastal Research
Vol. 27 • No. 6
November 2011
Vol. 27 • No. 6
November 2011
autoregressive
cyclical autoregressive models
Guyana coast
mudshoals
stochastic processes
Time-series modeling