Extended cellular automaton models of solar flare occurrence

In a previous study, we proposed a class of stochastic cellular automaton m odels for the occurrence of solar flares. 'Flaring elements' of a deliberat ely unspecified nature were allowed to interact. Investigation of a 1-D ver sion of the model, capable of analytical description, conflicted in detail with observations but indicated the qualitative success of such a picture. In this extended study, we investigate the consequences of relaxing various assumptions present in the initial study. In this way we can build up a fa mily of models each producing a better match with the observations. From th ese results we can say which features control the power-law indices obtaine d from self-organising models of flare occurrence. The sorts of constraints this type of modelling produces will also have to be satisfied by more phy sically detailed pictures.