On linear birth-and-death processes in a random environment (Q2510390)

The authors investigate the quenched probability \(p\) of extinction of single-type and multi-type birth-and-death processes with finitely many Markovian environments. In particular, they prove that, for a single-type process, \(p=1\) almost surely if and only if \(\sum_k (a_k-b_k)u_k\leq 0\), where \(a_k\) and \(b_k\) are the birth and death rates in the environment \(k\), and \(u_k\) is the average proportion of time spent in the environment \(k\). A similar result is also obtained for a multi-type process. The authors' findings are accompanied by a number of interesting and clarifying comments.