A family of stochastic models for droughts

Abstract

There is no single quantitative definition of a drought; what constitutes a drought depends both on water needs and on storage capabilities. In this paper a family of models is proposed which is sufficiently flexible to cover a range of different situations; the user may select some of the parameters to suit his particular application. The input to the model is a sequence of rainfall values which in turn may be generated by a stochastic model, thus enabling the simulation of a long sequence of rainfall data which allow the user to study the distribution of droughts of any given duration or severity. Several of illustrative examples are given.