Fatigue statistical distributions useful for modeling diameter and mortality of trees

Mortality processes and the distribution of the diameter at breast height (DBH) of trees are two important problems in forestry. Trees die due to several factors caused by stress according to a phenomenon similar to material fatigue. Specifically, the force (rate) of mortality of trees quickly increases at a first stage and then reaches a maximum. In that moment, this rate slowly decreases until stabilizing at a constant value in the long term establishing a second stage of such a rate. Birnbaum-Saunders (BS) distributions are models that have received considerable attention currently due to their interesting properties. BS models have their genesis from a problem of material fatigue and present a failure or hazard rate (equivalent to the force of mortality) that has the same behavior as that of the DBH of trees. Then, BS distributions have arguments that transform them into models that can be useful in forestry. In this paper, we present a methodology based on BS distributions associated with this forest thematic. To complete this study, we perform an application of five real DBH data sets (some of them unpublished) that provides statistical evidence in favor of the BS methodology in relation to the forestry standard methodology. This application provides valuable financial information that can be used for making decisions in forestry.