Lower deviations of branching processes in random environment with geometrical offspring distributions

We consider the branching processes in random environment. In this paper, we deal with the case of environments which are chosen stationary and ergodic from the finite set of geometrical offspring distributions. We denote by Z n the population at the n -th generation. We show that the large deviation principle holds with a certain rate function for the total population when the environment satisfies some conditions. Also, we will show that the trajectory t → log Z n t n , t ∈ [ 0 , 1 ] converges to a deterministic function uniformly in probability conditioned on { 0 < Z n ≤ e c n } .