Reconstructing the history of forest fire frequency – Identifying hazard rate change points using the Bayes ’ infor - mation criterion

A problem in model selection, namely the identification of multiple change points for a piece-wise constant hazard rate, is discussed. A methodology using the Bayes’ Information Criterion is developed in an overdispersed survival model (with corresponding quasi-likelihood function). The technique is used to identify changes in the historical frequency of forest fire. It is applied to two datasets derived from time-since-fire maps in the Canadian Rockies.