An Application of Markov Chain Analysis to the Ayrshire Coal Measures Succession

Abstract

The vertical ordering of rock types in a Coal Measures succession is analysed in terms of an ‘embedded Markov chain’ model. The presence of a statistically significant ‘memory’ within the sequence is proved and a Markov model of combined first and second order properties established. The form of the model is one of loosely repeating basic units (corresponding to cycles) whose internal structure is suggestive of origin in fluvial and floodplain environments and control by local vegetational and sedimentary factors. Prediction of the cycle member content distribution by various models indicates that the ‘cyclic’ behaviour of the succession is satisfactorily accounted for by the second order Markov model, although there is the possibility of an additional longer term control.