Complete left tail asymptotic for branching processes in random environments

TL;DR

This paper develops a new method to derive the complete left tail asymptotics for the density of the martingale limit in branching processes within random environments, overcoming previous functional equation limitations.

Contribution

It introduces a novel approach using ratios of probabilities of rare events and generalized Schr"oder-type equations to obtain asymptotics in complex random environment models.

Findings

Derived asymptotic series converging everywhere.

Established connections with fractal measures.

Extended classical tail asymptotics to random environments.

Abstract

Recently, the complete left tail asymptotic for the density of the {\it martingale limit} of the classical Galton-Watson process has been derived. The derivation is based on the properties of a special function (whose inverse Fourier transform is the density) satisfying a Poincar\'e-type functional equation. However, the corresponding special function does not satisfy the Poincar\'e-type functional equation for the processes in random environments. This makes it impossible to apply the developed asymptotic methods. In this instance, even computing the special function and the density becomes extremely challenging. We propose a method that extracts the terms of the asymptotic series from another limit distribution {\it ratios of probabilities of rare events}. The generating functions for this limit distribution satisfy the generalized Schr\"oder-type functional equations that greatly simplify the calculations. As in the classical case of constant environments, the obtained series converges everywhere. Possible interesting connections with exotic fractal measures that Robert Strichartz considers are also discussed.